strike that, I didn't lose it after all! Yay! :-)

With the new system, reload before wincing, cursing, or throwing the screen out the window.

I see that I have a "Live Preview" here.

I see that I can go directly to the end of a thread by selecting the number of messages (101) rather than the title (Monte Irvin).

I apologize for what I have done to Monte Irvin's thread.

>I see that I can go directly to the end of a thread by selecting the number of messages (101) rather than the title (Monte Irvin).

Wow, this is the greatest thing since sliced bread. (And about time too! ;-)

OK. I finally had a bit of free time this evening, so I've run MLEs for Irvin.

The results, as you will see, are pretty similar to Dr. Chaleeko's.

A few methodological differences.

1) I have used only NNL, MxL and MiL data. My general practice has been to leave out CWL play unless it constitutes a very significant part of a player's career (e.g. Oms), so I have done so in Irvin's case. I have never used PWL data, so I have omitted it here also.

2) I have used the following conversion factors: .90/.82 for the NNL, .87/.76 for the MxL, and .93/.86 for the IL

3) I have regressed Irvin's seasons through 1948. I did not regress 1949 and 1950. I treated those as part-seasons. My projections should be added to Irvin's ML play to get his totals for those two seasons.

4) I did not translate any minor-league play for Irvin after he had established himself in the majors. Race may well have been a factor in his being sent down to the minors again in 1956/57, but there's enough evidence that Irvin was declining that I'm not comfortable including them myself.

5) I have simply skipped over Irvin's war years, giving him no projections for 1943-44 and only a part-season for 1945. War credit can be assigned as each of us thinks best. Note that, since Irvin clearly missed peak seasons during the war, the omission of these seasons will contribute to a lower career OPS+ for Irvin than the one calculated by Dr. Chaleeko.

I think that's it. I'll post the numbers in a few minutes, as soon as I have them formatted decently.

Monte Irvin’s MLEs


Year Team EqG PA BB Hits TB BA OBP SA OPS+
1939 Newark 75 301 22 98 135 .351 .399 .484 136
1940 Newark 128 504 41 151 234 .327 .382 .506 143
1941 Newark 141 556 52 168 265 .332 .394 .526 157
1942 New/MxL 146 591 60 170 296 .321 .389 .558 175
1943 Military Service
1944 Military Service
1945 Newark 30 120 14 28 46 .267 .350 .431 116
1946 Newark 144 605 77 176 269 .333 .418 .510 162
1947 Newark 150 630 81 159 257 .290 .382 .468 125
1948 Newark 116 487 65 128 218 .304 .396 .517 146
1949 JC* 63 263 41 75 120 .340 .443 .541 162
1950 JC* 18 80 20 28 61 .465 .600 1.025 318
1011 4137 473 1182 1902 .322 .400 .519 153

ML career 764 2893 351 731 1187 .293 .383 .475 126
Combined 1775 7030 824 1913 3089 .308 .389 .498 142
</pre>
I figure, with war credit, Irvin would probably have about 8300 PA with an OPS+ of about 145.

P.S. The playing time projections are somewhat different from Dr. Chaleeko’s also. I went with his summer-league playing time only as the basis for the projections.

P.P.S. I don’t have win shares to go with these projections yet, and I probably won’t have them this week, but hopefully playing time and OPS+ will be enough for everyone to go on this year.

Thanks Chris. Still by no means a slam dunk, but needs to be moved up onto the ballot next time, I think.

The biggest difference that I see is Doc's MLE's give Irvin a 199 OPS+ in 1942, whereas Chris's give him only 175. Chris, is this because of the regression? And which do you think is more accurate. A big year gives Irvin a Medwick like peak with more career, i.e. top five on my ballot. Without it he probably slides down a few slots.

Thanks

The irony is that Chris has stated many times that his method smooths out peaks, and we know from experience that indeed almost all of the NeLers come out with low peaks and high career numbers, relative to one another.

Regression pushes you back towards the mean. Thus when you have a 142-211-183 sequence like Irvin's, based on a relatively small number of ABs, regression knocks the top off the 211 just as June through September will to the player who is batting .455 in May. It is therefore more correct for ironing out random fluctuations, and not presuming that e.g. Meyerle's 241 in 1871 or George Hall's 207 in 1876 actually represents a Ruthian peak.

Where it's unfair is if you have a succession of say 165s, with the total number of ABs being sufficient to represent a real ability level. Then, and only then, it unjustly flattens the peak, while maintaining the overall average OPS+.

In this case, the 1942 season is so out of balance with the career as a whole and the surrounding years that it's unlikely to represent a real peak, but simply a random fluctuation.

Having said that, Chris's MLEs are sufficiently impressive that I think Irvin is HOM-worthy, if only marginally so, and will be on my top 15 next year.

Where it's unfair is if you have a succession of say 165s, with the total number of ABs being sufficient to represent a real ability level. Then, and only then, it unjustly flattens the peak, while maintaining the overall average OPS+.

Does the regression method use neighboring seasons to smooth the curve? I thought I recalled that this was the case. One player (can't remember who) had a simply dreadful year in mid-career that was also pulling down neighboring seasons. If I recall correctly, the regression was tweaked for that player -- keeping the dreadful season, but lowering its effect on the neighboring seasons.

If its a multiple season regression, than a succession of 165s would stay as it is.

Anyhow, many thanks to Chris Cobb and Doctor Chaleeko for being so patient with us regarding these MLE's and handling our requests for explanations, etc.

Well, it is still true that NeLers end up with low peaks. It might be statistically valid but in the real world we in fact get dramatic highs and lows that do not tend to exist in NeL MLEs. This is especially true of WS but also of OPS+.

Acc. to this logic, if 18 games in JC were all that Irvin had in 1950 (at OPS+ 318), what should it regress to? 142? 152? 163? 211 would be too much?

Of course, not being a statistician, I have no idea what that would regress to if you extrapolated 150-160 games from 18. But I do know that if you take all of the NeL MLEs and compare them to the real records of HoM caliber players, they make up two different clusters and the twain does not meet.

The fact that Irvin still has a couple 160s and a 170 in there says a lot about Irvin. The fact that he never gets above 30 WS says a lot about the method.

Since you regress proportionately to the square root of games played, and 18 is 1/9 of 162, then Irvin's 1950, regressed to a 162 game season based on his career OPS+ of 143 would be 201. Regression in that case is a bit too liberal, IMHO; do you really think 18 games at 318 is equivalent to a season at 201?

The fact that Irvin still has a couple 160s and a 170 in there says a lot about Irvin. The fact that he never gets above 30 WS says a lot about the method.

Which method, Marc? The version of WS that Eric is using or Win Shares themselves?

On 1942: Regression is not actually the decisive factor in my projecting Irvin at a 175 OPS+ vs. Dr. Chaleeko's 1999 projection. I get unregressed figures along the way in my analysis process, and using them to calculate an OPS+ gives Irvin a 176 OPS+ for that season. Since we have 80 games for Irvin that year and since his surrounding play is not far below this 176 OPS+ level, regression analysis suggests that over another 60 games, Irvin's regression towards his own mean would not have lowered his level of play for the season much at all.

I think the difference in our assessments arises from something in our handling of the MxL data. It's quite possible that I made a mistake in handling the MxL league data, which Dr. Chaleeko compiled himself.

I would note that if Irvin's 1942 is projected at 175 OPS+ rather than 199, that hardly shows him as lacking a Medwickian peak. Medwick's 1937 season was only 180 OPS+, after all.

Here's Medwick's 5-year OPS+ peak: 151, 156, 180, 140, 131, for a 152 OPS+ over that stretch
Here's Irvin's, dropping his brief 1945 season: 143, 157, 175, 162, 125, for a 152 OPS+ over that stretch.

Medwick was more durable during his peak, playing in 761 games to Irvin's projected 709, but there's little to suggest that Medwick's peak was substantially better than Irvin's because he carried a 180 OPS+ in his top season rather than a 175 OPS+.

There's several likely areas for variation between my unregressed numbers and Chris's for 1942.

1) The data source: I'm using
Veracruz MxL .397/.772 in 63 G and 287 AB
Newark NNL .531/1.156 in 8 G and 32 AB.

I think Chris is using the same, but if he was choosing to leave out the NNL data that would change things. I'm adding together conversions to create a full season from them since that's all of his summer play for the year.

2) The conversion rate: To get the 199 OPS+ figure, I converted the numbers like this:

MxL: .875/.76
NNL: .90/.82

If Chris chose different conversion factors that would change things, though I think those are his prefered rates for these leagues.

3) The walk rate: This one's where things get complicated. There's a number of possible places for difference.
-First off, there's the treatment of his Mexican walk rate. My method is to do a straight translation of walking from one league to the next, like this:

walk rate / lg walk rate * NL walk rate * playing time
</pre>
That gets me to Irvin walking 59 times in an NL atmosphere based on his Mexican rates projected into the NL.

-Now there's another little catch here: how do we each calculate walk rate? I do something perhaps unusual, but I calculate it as walks/hits. There's a few reasons for this but it boils down to this: you've got to have a constant to work against, but when traveling between so many fluid contexts, using hits instead of ABs avoids situations where applying a standard walk rate against a non-outcome based constant like ABs causes the walks to rise or fall too much. For instance, if I used walks/AB, the proportion of his RC that would come from walks would be increased simply by translating him from the Mexican League to the NL because the NL hit for a lower average than the MxL did (thus, fewer hits for Irvin since I key averages on league averages). Therefore, I used walks/hits so that the walks rise and fall in proportion to other offensive events. If Chris chooses a different measurement of walks, he may arrive at different results. (By the by, this is a nod to James's treatment of Willie Davis and Sam Crawford in the NHBA.)

-Also, while talking about the walks, my method takes a player's walk rate and compares it to his league. (I think Chris has the same walks info I do for the MxL in 1942, but if not, that's a place for difference.) I then place the player into his new context by multiplying the ratio of his walk rate to his league's by the walk rate of the NL (or whatever league I'm projecting him into). If Chris has a different routine for doing this, he will have different results.

-For the 1942 NgL season, I "created" Irvin's walks by using the same ratio of his walks to his league's walks that he showed in Mexico. I chose to do this rather than use career norms because it was from the same year and other walk data was from a different portion of his career. (The ratio of his career median walk rate versus the NL's was 1.23.). If Chris chose a different route to get there, he may have arrived at a different answer.

4) I just added the two sets of projections together to get a single 1942 line.

5) I could have been calculating OPS differently: I use what I believe is the David Foss method, taking care (I hope!) to use pitcher-removed league batting values. .328/.356 is the league context I'm using for 1942 per the SBE.

6) No park adjustments on either end of the translation. If Chris were putting Irvin into the Polo Grounds, that would be a difference, but I think he's putting Irvin into a neutral context.

Anyway, that's what I can think of off the top of my head; Chris what do you think?

Given that I have Irvin at 60 bb in 1942 and you have him at 59, I doubt that much of the difference comes from walk rates.

My suspicion is that I used the wrong baseline for calculating the offensive context. Dr. Chaleeko, do your MxL league avg./league slg. totals include pitchers, or not?

I took them as including pitchers and compared them to NL totals, not pitcher-removed totals, to match contexts. If the MxL totals do not include pitchers, then I should have compared them to NL pitcher-removed league averages, which would reduce the offensive conversion factors for the 1942 MxL considerably.

>Medwick was more durable during his peak, playing in 761 games to Irvin's projected 709, but there's little to suggest that Medwick's peak was substantially better than Irvin's because he carried a 180 OPS+ in his top season rather than a 175 OPS+.

Chris, I don't disagree with that but the point is that it feels like Irvin should have a higher peak than that (or than Medwick), not just equal to Medwick. I say that perhaps naively based on his 145ish career OPS+ versus Ducky's 133. I also meant to say that Ducky's 40 WS peak versus Irvin's 30 seems out of line, too, but of course I know you think it's the 40 that is too high whereas I find the 30 to be pretty low given everything else we think we know about Monte.

Anyway my real point was that at a macro level (and since macro I don't have any specific data points to offer) the NeLers seem to lack the high peaks of the better MLers, while the NeLers often seem to have fairly long careers given their other numbers. That's pretty impressionistic, but I'm sticking with it.

I mean that constructively as I hope you understand. I think you should author a book with all of these MLEs in them in order to help other baseball fans to sort out the mythology of the NeLers and especially to help them discover some of the gems that have thus far been overlooked by the general public (including Cooperstown).

Chris, I'm sorry, I should have been more clear when I sent those items to you. The MxL contexts are (mostly) pitcher-removed. I think Bragana or McHenry who would occasionally do duty in the OF are probably included, but I have been treating them as pitcher-removed. Cisneros, like Total Baseball, only gives hitting stats for pitchers when they have also played at another position.

Any Cuban leaguewide data is also, essentially, pitcher-removed.

I've assembled a lot more data since that last one I sent you, I'll send the new one to you.

Chris, I don't disagree with that but the point is that it feels like Irvin should have a higher peak than that (or than Medwick), not just equal to Medwick. I say that perhaps naively based on his 145ish career OPS+ versus Ducky's 133. I also meant to say that Ducky's 40 WS peak versus Irvin's 30 seems out of line, too, but of course I know you think it's the 40 that is too high whereas I find the 30 to be pretty low given everything else we think we know about Monte.

The height of Irvin's peak in win shares may be different in my projections than in Dr. C's, because I am projecting Irvin for somewhat more games played in some of his top seasons, and I am planning to model a typical career curve for his fielding value, whereas I think Dr. C has used a flat career rate for fws.

However, if Irvin ends up with the same OPS+ in his best season as Medwick, he will probably have a lower win-share total, because Medwick's bws/OPS+ ratio is among the top 10 over a 50-year period for players with seasons of 175-185 OPS+. The ws-estimate method necessarily aims at the average ws value for a given OPS+. In this respect, the projection method certainly smooths out player's careers in comparison to Jamesian win shares, steering a middle path, as it were, between Joe Medwick and Bob Johnson. I don't think there's any viable alternative to this smoothing process.

It's also true that in many cases, regression also smooths out peaks and values somewhat, since regression to the mean is only showing the most probable outcome, and other outcomes would happen in reality. Since we can't guess when exactly less probable outcomes would have occurred, I don't see a viable alternative.

This whole smoothing business is less of an issue for me than for peak voters who include a discontinuous peak measure in their systems, since I look only at groups of seasons. I would suggest that peak voters who include this element might note the extent to which high single-season peaks (esp. in win shares) can be influenced by team context, team luck, and player luck.

Chris, I'm sorry, I should have been more clear when I sent those items to you. The MxL contexts are (mostly) pitcher-removed.

Thanks for confirming my suspicion. I'll recalculate Irvin's career using a corrected offensive context factor for 1942. I'll also add in a context factor for his IL seasons, which I didn't have before.

Irvin's 1942 season will certainly be rising in the translation, which will have some small impact on surrounding seasons as well.

I won't know the effets on 1949 and 1950 until I see the league data.

New Irvin totals from me may not be posted until tomorrow night, but I hope to have win-share estimates to go along with them.

New Irvin totals from me may not be posted until tomorrow night, but I hope to have win-share estimates to go along with them.

I was going to post my ballot tonight, but I'll hold off on doing that now.

Will Clark career (8283 PA): 303/384/497, OPS+ 138
Monte Irvin career(7030 PA): 308/389/498, OPS+ 142
(per Chris’s MLE’s)

Will Clark’s top nine season OPS+, from best to worse (minimum 400 PA)

175 60 53 52 50 45 40 28 27 26

Monte Irvin’s top nine season OPS+, from best to worst (minimum 400 PA)
(per Chris’s MLEs)

175 62 59 57 47* 46 42 25 25 20

(*Dr. C’s unregressed figure for the 1949 season)

As I mentioned before, Clark and Irvin both had problems staying in the lineup after age 30, and were out of baseball by age 37 (Clark) or 38 (Irvin). Irvin earned 96 WS after age 30 in the majors (he would get a couple more for his ’51 season); Clark earned 90.

The rate stats for Clark and Irvin are pretty close. Of course, Irvin is missing two prime years to the war; having that information would change the comparison considerably. But I still thought this was interesting.

Monte Irvin’s MLEs


Year Team EqG PA BB Hits TB BA OBP SA OPS+ BWS FWS Total
1939 Newark 75 301 22 98 135 .351 .399 .484 136 9.9 1.7 11.6
1940 Newark 128 504 41 152 235 .327 .384 .506 144 18.9 3.1 22.0
1941 Newark 141 556 52 169 267 .332 .396 .526 159 24.0 3.9 27.9
1942 New/MxL 146 591 60 174 301 .321 .395 .558 180 32.0 4.8 36.8
1943 Military Service
1944 Military Service
1945 Newark 30 120 14 29 46 .267 .352 .431 117 3.1 0.7 3.8
1946 Newark 144 605 77 176 269 .333 .419 .510 163 28.6 3.8 32.4
1947 Newark 150 630 81 159 256 .290 .382 .468 124 16.7 4.0 20.7
1948 Newark 116 487 65 128 219 .304 .396 .517 146 17.0 2.4 19.6
1949 JC* 63 263 41 75 120 .340 .443 .541 162 11.7 1.0 12.7
1950 JC* 18 80 20 28 61 .465 .600 1.025 318 5.1 0.5 5.6
1011 4137 473 1188 1910 .322 .402 .519 154 167.0 26.2 193.2

ML career 764 2893 351 731 1187 .293 .383 .475 126 83.3 14.0 97.3
Combined 1775 7030 824 1919 3097 .309 .390 .499 143 250.3 40.2 290.5
</pre>

I figure, with war credit, Irvin would probably have about 8600 PA with an OPS+ of about 146. I’d peg him at about 363 career win shares with war credit.

Here are Irvin’s updated MLEs, with projected win shares.

Notes.

The main change from my previous projections is for 1942, as I expected. My previous translation was too low. I had misidentified the pitchers-excluded batting and slugging averages for the MxL that year as pitcher-included, so I was overestimating the offensive context by about 5%. Because of the regression, the change in this season also shows up in small changes in surrounding seasons. The unregressed OPS+ for Irvin’s 1942 is 184, so my estimate is now much closer to Dr. Chaleeko’s projection of 199 for this year, but we’re still differing somewhat. I do not currently have a hypothesis about the source of that difference, but I’m going to continue to look for the source. The change of the 1942 OPS+ estimate from 175 to 180 and its effects on surrounding seasons raise Irvin’s career OPS+ by 1, from 142 to 143.

Irvin’s fielding win shares are based on the assumption that in the majors he would have been a corner outfielder for his whole career. To get an estimate of how good he might have been defensively prior to age 30, I looked at two contemporary examples of outfielders of considerable talent slowed by injuries in their 30s, Joe Medwick and Joe Dimaggio. Medwick’s fielding rate from 30 on was 82% of his pre-30 rate; Dimaggio’s was 84%. Irvin was a C+ outfielder in the majors. No exact rate was available in Win Shares, so I took the middle of the C+ range, 2.6/1000, and took that as 86% of Irvin’s pre-30 fielding rate. That came out to about 3.0/1000, so I projected a curve of seasonal rates with a peak year of 3.8/1000 that averaged out to 3.0/1000.

Irvin’s batting win shares are estimated by my usual method of finding the closest OPS+ matches, adjusting for OPS+ differences, and projecting the estimated rate into Irvin’s projected PAs. His career bws/PA is .03560. Other pre-DH players with a 143 OPS+ include Harmon Killebrew (.03429), Lefty O’Doul (.03635), and Eddie Mathews (.03835). Irvin’s values fit within this group, so this estimate is reasonable.

Chris, I think the difference is in our walk estimates. I estimated Irvin with 59 MLE walks for his Mexican play only, then another 20 for his NNL play for 79 total. Sorry if my previous post to that effect didn't make that clear. If I took him down to 60 walks, then his OPS+ would be 189, within 5 points of your estimate.

Per Chris' WS estimates, it would seem that for a peak voter Irvin's case may depend upon how much credit he is given for the war. His three best years are the two immediately before and the only immediately after the war. Give him three seasons of 30 WS (maybe 26 or 27 in 1945 to bring th total up to 30) and he looks pretty good. It that would also give him about 380 for his career so maybe I am a bit high on that estimate.

I give credit to Irvin for 400 games missed during the war, at a rate that is 90% of his per game value in the surrounding seasons of 1941-42, 45-47. Without the cautionary 10% deduction, that's 27.2 win shares for two 140-game seasons in 1943 and 1944 and 23.3 additional win shares for 1945. He's better than that in 1941, 42, and 46, but his lesser 1947 brings his rate down a bit.

My own adjWS for Irvin would look like this:

334 37-32-28/140

That is pretty conservative, as you can see, but definitely ballot-worthy.

http://www.nj.com/search/index.ssf?/base/news-1/11339383268250.xml?starledger?nex&coll=1

In case the link doesn't work, it's a Newark Star-Ledger story saying that he may have a park in his native Orange, NJ named after him.


"It's always been a wonderful park with its floral arrangements and cherry trees. It had two baseball fields, and on Sunday you'd stroll through there with your best girl in your Sunday best," reminisced Irvin, 86, who now lives in Homosassa, Fla. "We really grew up in Orange Park."

Note: I have not recently read any of this "Monte Irvin" thread.

Recently 'burniswright' asserted simply that Monte Irvin was not a CF. KJOK guessed that he means not in the major leagues.

John Holway seems to make Irvin a shortstop more than anything else. See the information relayed by Chris Cobb in the CF discussion


Earlier in that thread I wrote,
20. Paul Wendt Posted: August 25, 2008 at 12:05 AM (#2915112)
In the Biographical Encyclopedia, Jim Riley lists 'cf ss 3b' for Monte Irvin, with none of the three bold (primary) or italic (emergency). That means cf first among equals, no primary position, although there are some mistakes in execution (bold or italic markup omitted; eg, Roy Campanella 'c 3b of p but the article treats him as a primary catcher).

It won't hurt me to read the Riley prose on Irvin, 2 full pages (second ed., 2002, p407-09)

. . .
A versatile ballplayer, he played both infield and outfield with the Eagles ['37-'42, '45-'48], often starting at third base or shortstop but eventually finding his niche in center field. . . .
[Mexico 1942, no comment on fielding]
. . . owner Effa Manley demanded compensation and the Dodgers withdrew their claim, allowing the New York Giants to seize the opportunity to sign the star outfielder. . . . the hard-hitting outfielder was assigned to Jersey City in the International League in 1949 . . .

If Holway is right that Irvin was simply a regular shortstop for two seasons after the war, it appears that Riley doesn't missed it.
"often starting at third base or shortstop" --is it plausible that that means he moved from infield in the spring to a summer niche in the outfield, repeatedly?

Ranking the Center Fielders #49 - following posts may be worth checking

49. Chris Cobb Posted: August 25, 2008 at 05:54 PM (#2915766)
Holway's positional listings back up burniswright's claim that Irvin was not a centerfielder, really. Here's what Holway has:

1939 -- RF
1940 -- CF
1941 -- SS
1942 -- ?? (Mexico)
1943-45 military service
1946 -- SS
1947 -- SS
1948 -- LF

I agree that Irvin probably should have been listed in left field. It would be very interesting to know how many games he really played at shortstop, and what position he played in Mexico (the MxL encyclopedia doesn't say).

Thanks, Paul, for spotting the evidence that probably led to Irvin's CF placement!

An additional note, in 1941, Holway has Leon Day in CF. I recall a story (maybe it is recounted on the Leon Day thread?) about how there was an injury that caused Day to take over as regular center fielder, so Irvin's play at SS that year seems to have been provoked by circumstances.

Holway doesn't give any sense, though, that Irvin was not a full-time SS in 1946-47, so he and Riley appear to be telling conflicting stories here.

Jim Riley once commented that Irvin played shortstop in Newark because on a team that lacks a shortstop, you put your best athlete there. I know that sounds a lot like high school, but you have to put it in its proper economic and socio-cultural context in order to make it sound less condescending.

One of the funk factors in NeL history is that whereas the established franchises were pretty responsible about putting guys on the field where they were supposed to be, defections, injuries, the draft, Mexican money, and any number of other factors could and did monkey-wrench those plans from time to time. And of course NeL teams didn't have AAA and AA guys waiting to move up. Irvin played CF after Fran Matthews, and before Jimmy Wilkes, who were both legit CFs. So Newark had a gap for portions of a couple years, and Irvin filled that also.

As to Irvin starting out as a shortstop, also consider that in the NeL environment, without the benefit of extensive scouting of players at lower levels, when a talented kid comes up it may take you awhile to figure out what to do with him. I was always astounded to find that when Martin Dihigo first showed up on Alex Pompez's Cuban Stars, they used him at 1B. What a waste! But the gangly kid was 6'2", and could hit the ball a mile. Ergo: a firstbaseman. Eventually they figured out that he was better used at 2B, 3B, CF and P. In fact, everywhere on the diamond but C and 1B. But it took awhile.

Cross posting from the CF thread RE: Monte Irvin

I think we need to figure out what position he should be ranked at (sounds like LF to me based on what I read above), and ask the voters in that election to slot Irvin and re-tally the votes. We need to get it right, even if it's slightly painful.

That doesn't mean everyone has to submit a new ballot. Just let us know where you'd put Irvin had he been eligible. Ballot counters, is it easy to just slot in a new player? Or would it be necessary to have everyone re-list each player?

Experts - how big of an error is it to list him as a CF? Is there a strong case that he would have played CF 1942-45, for example?

Does anyone have access to a newspaper that covered the Newark Eagles.
For example,
where did he play during the first weeks of 1942 before disputing to Mexico? There would be only a few league games to look up. And pre-season? Did the Giants, Yankees, or Dodgers play exhibition games vs the Eagles including Irvin?

Can I get updated year-by-year breakdowns of MLE games played at each position and total Fielding Win Shares? I'll need them to do WARP.

136. Paul Wendt Posted: August 27, 2008 at 09:46 AM (#2918408)
Does anyone have access to a newspaper that covered the Newark Eagles.
For example, where did he play during the first weeks of 1942 before disputing to Mexico? There would be only a few league games to look up. And pre-season? Did the Giants, Yankees, or Dodgers play exhibition games vs the Eagles including Irvin?

Finally, a chance to add something to your research! Here is a summation of every pertinent mention of Monte Irvin from 1942 in the Pittsburgh Courier, Chicago Defender and Philadelphia Tribune. I also searched the New York Times and Los Angeles Times, neither of which ever had a mention of Irvin.

The NYT had a line score (including batteries), but no boxscores of the 5/31 Ruppert Memorial Cup at Yankee Stadium, which included the Baltimore Elite Giants, Philadelphia Stars, Newark Eagles and New York Cuban Stars. Even a brief story on the Negro All-Star game in Chicago included only a line score and batteries. The only other mentions of Newark Eagles games were from July in the Metropolitan Baseball Association in the "Events Today" box, but never a story on any of the games.

From the three black papers:

Pittsburgh Courier 3/14/42: Irvin not back in states from Puerto Rico because of ship delays due to German U-Boat activity in the Caribbean. Air travel from PR denied for civilian purposes.

Chicago Defender 4/18/42: Preview of 4/19 Newark Eagles v Homestead Grays exhibition at DC. "On . . . Sunday, May 3, they [Homestead Grays] meet their arch enemies, Newark Eagles for the third time in four weeks. This will be the official opening of the Manley crew on their home grounds . . ."

PC 4/23/42 and CD 4/25/42: NE v HG exhibition at DC (Griffith Stadium, 4/19/42), doubleheader, but only first game box score given, with Irvin in CF. Batting Order: Israel 3B, Wells SS, Stone RF, Irvin CF, Dandridge 2B, Clarkson LF, Pearson 1B, Parks C, Hobgood P (subs: Hill RP, Day 1B)

PC 5/16/42: New York Cubans at NE (5/10/42, Ruppert Stadium), Negro National League doubleheader, no boxscore. The only mention of Irvin's play in both games, aside from his hitting: "He also made an unusual throw for a double play." In a separate story: NE will play two games at Philadelphia Stars on 5/17-18. No mention of Irvin (or anyone else.)

PC 8/1/42: Report that Irvin is among highest paid players in Mexican League. In a separate story, Chester L. Washington: "In a response to a volley of inquiries, here are a few sepia players who stand a good chance of making the grade if given tryouts in the major leagues: EAST--Catchers, Josh Gibson and Campanella; pitchers, Leon Day, Partlow and Gaines; infielders, Easterling, Hughes and Blanco of the Cubans; outfielders, Benjamin, Bill Wright, and Irvin (now in Mexico). WEST--Infielder Strong and Outfielder Brown.

PC 11/7/42: Cum Posey's all-NgL team for 1942 (19th annual, starting 1924): "The year of 1942 was the pinnacle of colored baseball, East and West." "We regret that Monte Irvin deserted Newark after playing one league game. Irvin, in our opinion, is the best player in present-day Negro baseball." Posey's 1942 team: P(2) Leon Day and Satchel Paige, C Josh Gibson, 1B Leonard Pearson, 2B William Serrell, 3B Howard Easterling, SS Willie Wells, LF Willard Brown, CF Jerry Benjamin, RF Dan Wilson, Mgr Frank Duncan.

We just did LF. I say we simply do a 1-week re-vote so that Irvin can be properly slotted.

There is no reason to slot Irvin in LF quickly. Get more response here unless the resolution no longer depends on the facts of his NeL career because he played "too much" LF in the majors.
(BTW there is a problem with career outfielding records at baseball-reference, which I have reported to Sean Forman.)
--

On a library PC this afternoon I read several articles in the Pittsburgh Courier and Chicago Defender, mainly 1940-42. The text file of notes that I made and saved there is corrupt on the flash drive here at home. I'll return tomorrow and plead for a rescue. That's all for now.

djrelays:
Here is a summation of every pertinent mention of Monte Irvin from 1942 in the Pittsburgh Courier, Chicago Defender and Philadelphia Tribune.

Did you run one search, 'Monte Irvin'?
Is Philadelphia Tribune a black daily or weekly that you can access via ProQuest.com?

Paul--

In all five papers I searched "Monte Irvin", "Irvin AND Newark", and "Newark Eagles" for the dates 1/1/1942 to 1/1/1943.

Yes, Philly Tribune is a black daily (no Sundays) that I'm pulling through ProQuest.

djrelays,
Thanks. When I go to the library I have Pittsburgh (new) and Chicago, not Philadelphia.
--

Here are my notes on Monte Irvin. I searched for "Monte Irvin" because I was looking for pre-season, opening day, post-season prose --not looking for box scores that others have compiled. I searched beginning 1937 and found nothing until 1940, then some interesting items. This helps know Irvin's career and know Negro Leagues baseball.

bold marks items that pertain to Irvin's fielding positions
cd = Chicago Defender
pc = Pittsburgh Courier



1940

0504 cd [That is, 1940-05-04 Chicago Defender]
Three new faces, inclg Clarence Chreal [Israel] to start at 3b. "He will take over Monte Irvin's former assignment. Irvin will be shifted to cf where Orange youth is expected to bolster the Eagles considerably."
1939 rookie Vernon Riddick is expected to start at shortstop, hand injury in training camp, expected to be ready.

0518 pc (pre-home opener)
Eagles infield includes new faces in three position . . .
Monte Irvin ... who was at third and sometimes in center field last year, is a shortstop. Original plans to have Irvin in center this year were altered for the time being."

0520 cd
opening doubleheader
split 6-1 and 3-4. one boxscore, Irvin 6 ss

0629 cd
double-header in the first-half-closing series
Israel dnp
"Monte Irvin turned in a brilliant performance at his old position, 3b, with a series of unusual throws. ...
Bus Clarkson, former WIlberforce athlete and Spoon Carter ... were new faces in the combine."
Clarkson ss and Carter p acquired by the Manleys in exchange for rights to Satchel Paige


1941

0412 cd
Leon Day has rejoined the Eagles, "among the first of the outlaw players to sign.
The NNL recently adopted a rule by which they gave outlaw players an opportunity to sign before May 1."
Day jumped the team and played in South America last year.

0503 cd
Late arrival Lenny Pearson is working out at 3b and outfield.
He and Irvin cf and Clarkson ss played in the Puerto Rican winter league playoffs.
[or Irvin not in the playoffs, not late -0412]

0517 cd
opening game Irvin cf

0621 pc by CUm Posey
"Monte Irvin the Newark Eagles infielder" is in a close batting race

0802 pc
0802 cd
BOth newspapers printed the same article on the exhibition season in Newark
dateline Newark 0731. These are non-league activities while the East-West classic
"took Monte Irvin ss, Lenny Pearson 3b, and Jimmie Hill p from Newark"

1025 pc
by Cum Posey, 18th annual All-America team
FIRST TEAM
. . .
IRVIN (Newark), lf
WRIGHT (Mexico), cf
STRONG (Kansas City), rf

1942

0131 pc by Cum Posey
"Monte Irvin of the Newark Eagles led the N.N.L. in 1942 [sic] in batting.
Irvin was the star of the year in Negro baseball."

0314 pc by Robert Hughey
Diamond Stars In Puerto Rico May Be Victims Of Axis Blockade
Possible BLockade Owners' Problem
Outstanding Players of Both Leagues Now on Islands as War Gets Nearer Home

Axis submarine threat to surface travel.
"the Puerto Rican government will not issue plane reservations for civilian purposes"
Terry McDuffie and Bill Byrd returned a few weeks ago.
long list by NeL team, including Monte Irvin Newark


1107 pc by Cum Posey
19th annual All-AMerica team
[no discussion or mention of any player outside the NNL and NAL afaik]
... "We regret that Monte Irvin deserted Newark after playing one league game.
Irvin, in our opinion, is the best players in present-day Negro baseball."

1943

0403 pc by Cum Posey
the all-time all-American club
after listing numerous players by position
"You have Dihigo, Butt, Bankhead, Cannaday [newline without punctuation or conjunction]
Monte Irvin to play in any position and nine hundred ninety nine others."

0925 pc
by Cum Posey

1944

1945

1946
0316 cd by Al Monroe, Swinging the News (not a baseball column)

"The Brooklyn Dodgers may add a third Negro baseball star, Monte Irvin, third baseman,
who hit a potent .365 with San Juan in the Puerto Rican League this winter.
He was voted M.V.P. in that association."

from the New York Amsterdam News, ignoring what is redundant

1939

0729 nyan
"Manager Lundy is experimenting with a series of combinations which promises to
bring the Eagles back home, the strongest they have been this season.
The principal switching is at third base where LUndy is trying to use Vernon Riddick,
recently obtained utility man.
Monte Irvin, New Jersey schoolboy who is rapidly becoming a standout in the N.N.L.
is the feature of the change. Irvin hits exceptionally well and Lundy is trying him
in centerfield with Lenny Pearson in leftfield."

0819 nyan
Newark Eagles Set For Pennant Chase
"Vernon Reddick has won recognition in the Eagle infield during the current road trip
and Monte Irvin was shifted to an outfield berth."

1940

0511 nyan by Dan Burley
"Negro baseball's big show opened here Sunday
. . . Israel was on third; Monte Irvin, the former third sacker, was at shortstop;"
box Irvin 2 ss
[That was the opening game at Ruppert Stadium 0505 but not a league game.]

0720 nyan
Eagles Revise Their Lineup
home series opens Thursday night, Israel 2b, Pearson 3b
"Israel ... started the season here at 3b. Pearson ... has been in CF most of the year.
Monte Irvin of Orange, former hot corner ace, will go to the field."

"The Eagles finished in third place in the first half ..."
[maybe this concerns the second half opener?]

0727 nyan
3-game sweep
Israel 3b "has been shifted to 2b"
"While Israel was recuperating Monte Irvin, now in CF, and Lenny Pearson, then the CF,
was alternating at the hot corner. Pearson ... won the assignment."

1941

0802 nyan by Dan Burley Sports Editor Amsterdam StarNewsSpecial
Biggest Crowd in History (50,000) Sees East Beat West, 8-3
photos courtesy Chicago Herald-American
Irvin 6 3b, two hits

1942

0425 nyan
Willie Wells Newark Boss
opening day just 10 days away
"Monte Irvin, who led the league in batting last year, is with the team in training
camp at Richmond VA ... quartered at Virginia Union College ...
Irvin, married recently, is eager to make a mark for himself in baseball. He
will play in the coveted centerfield spot with Ed Stone flanking him in leftfield."

0606 nyan
action photos show
Monte Irvin, Newark (no date) [is he really in Mexico?]
Ray Dandridge, Newark infielder, surrounded on the ground after beaned by Barnhill Sunday
Barnhill safe at third "as Newark's Walker gets the ball but way off base."
[maybe Dandridge 3b and Walker his replacement after beaning]

0725 nyan by Dan Burley
Here's a Line on Top Candidates for Major Loop Consideration
"Irvin Prize Catch
Outfielders would include first of all, the sensational Monte Irvin, Newark's great outfielding shark who has one of the best throwing arms in Negro baseball and who can hit with the best of 'em."
[is he really in Mexico?]

1943
1944
1945
1946

1947

0426 nyan
Polo Grounds Sunday 0427, Eagles-Cubans
Eagles trained at Jacksonville, Cubans at New Orleans
Cubans 6-0 perfect game v New Orleans Creoles
Eagle lineup is set inclg lf Davis, cf Wilkes, rf Irvin, 1b Pearson

0503 nyan
box for that game shows the lineup not set after all
lf Davis, cf Irvin, rf Harvey, 1b Pearson
- Wilkes, substitute cf, and Mgr Mackey pinch-hit for Wilkes in 9th

Here is the link for the file you uploaded:
http://www.yousendit.com/download/Q01FWGJBaFJKV00wTVE9PQ

That pdf file includes three articles by Cum Posey where he picks so-called all-America teams, fall 1941, fall 1942, and spring 1943 about the "all-time all-America club" --maybe under discussion or selection elsewhere?

I agree with Paul - there is no rush, let's get all of our ducks in a row first . . .

Adding to Paul's Cum Posey All-America Teams, he also picked one in December 1937, which Gary has on his website:

Gary Ashwill Agate Type Website

That one is another "all-timer". The fall 1941 and fall 1942 articles are called the 18th and 19th annual.

I have uploaded the file to Baseball-Fever where it will be available after #144 expires today.
Posey all-America teams

Hey guys, I've been away for awhile, I've got a major test to study for the next few weeks (I take it October 4).

Just checking going up from bottom to top on the bar on the right side . . . is there anything left to do to add Monte Irvin to the LF ballot? Are we in agreement that that is best?

Any suggestions on how to go about slotting Monte Irvin for LF?

Should we just add the notes to this thread, or the LF ballot thread?

I would suggest a new "Add Monte Irvin to the LF Ballot" thread. Ask everyone who voted in the LF election to state where they would have put Irvin on their ballots. Ideally, we would get everyone who actually cast a LF ballot - I do have records of their votes and I think I can figure out how to count it - if no one changes anything else.

Monte Irvin MLEs, second edition



1942* Veracruz in MXL; Also played for Newark
1945* Missed time in this season is due to military service; credit appropriate
1949* Jersey City in International League; also played for New York Giants
1950* Also played for Jersey City

Notes on conversions.
For 1942 MxL, conversion factors of .87/.76 were applied
For 1949-50 IL conversion factors of .93/.86 were applied
For 1939-48, regression to a 3-5 year mean was applied.
For 1949-50, minor league stats were converted to MLEs and added to major-league stats for those seasons to produce seasonal totals.
For 1951-56, totals are for major-league play only

I haven’t done estimated extra base hit totals. Since Irvin has a major-league profile, I assume that will provide the necessary information for estimated baserunning wins for his pre-ML seasons. His higher stolen base totals before WWII would very likely have been matched by higher totals of triples in those seasons. I would guess his base-running value was pretty consistent 1946-51; after that he was significantly hampered by ankle injuries.

Positions played?

Oh, and Fielding WS, of course.

Fielding win shares for Irvin, based on the same playing time estimates that appear in these MLEs, are available in post 123 above, along with an explanation of the basis of the estimate.

That post assumes, however, that Irvin would have been a corner outfielder throughout his career in the majors. If you want fielding projections that move him around through the various positions he played in the Negro Leagues, I could work those up, but you'll be waiting another week!

Since we're ranking Irvin as a left fielder, I'd suggest we project him as a left fielder.

Yes, I definitely would like projections based on the positions he actually played in the Negro Leagues, for the sake of consistency with other NgL'ers (Beckwith is credited for his time at SS, etc.). I can wait as long as you need to produce them--what's the rush?

Yes, I definitely would like projections based on the positions he actually played in the Negro Leagues, for the sake of consistency with other NgL'ers (Beckwith is credited for his time at SS, etc.). I can wait as long as you need to produce them--what's the rush?

I have to agree with Dan for consistency's sake, too.

Overall, I have projected Irvin as a competent fielder at all of the positions that he played, but without a strong fielding peak anywhere. There were few seasons in his career in which he appears to have played at only one position, and in the course of his career he started for some stretch at every position except pitcher and catcher. This suggests that he was a highly talented athlete, but it points toward a jack-of-all-trades fielding profile. Table 1 below gives the fielding win share totals. Table 2 gives seasonal breakdowns of defensive games played by position, along with fielding rates at each position. Tables 3 and 4 give career game totals for non-ML play and for total career. Finally, I comment on the basis for my decisions about Irvin’s positions for each season.



Comments on Positional Assignments, Season by Season
1939. Newspaper accounts suggest manager Lundy started Irvin at 3B but moved him to CF around the beginning of August. In the NeL season, Irvin probably played 2/3 of his games at 3B, but the MLE projected playing time shows a mid-season call-up, so I give him essentially 1 major-league month at 3B, and 2 in center field.
1940. Newspaper accounts suggest the goal is again to play Irvin in centerfield, but he starts the season at shortstop and plays there until Newark acquires Bus Clarkson, who takes over at short. If a more precise date for this transaction were found, I could fix the percentage of games played more exactly. Irvin played some at third base, but Lenny Pearson was preferred at that position.
1941. Clarkson jumps to Mexico, so SS is again a problem. A replacement is found, but succumbs to injury. Leon Day becomes the everyday CF, and Irvin covers SS. He is never viewed as the solution at the SS position, but his ability to cover the position keeps being called for.
1942. Played CF in exhibition games and at the very beginning of NeL season. Dr. Chaleeko’s post of MxL data for that year lists Irvin at 2B. I don’t know his source, but that is the only piece of information I have about Irvin’s MxL position, so I’ve gone with it.
1945. Negro Leagues book that was Dr. Chaleeko’s data source lists OF; I have assumed center since that is the OF position Irvin had consistently played.
1946-47. Negro Leagues book lists SS/OF and OF/SS for these seasons; I have gone with a mix that shows a transition away from SS/CF to corner outfield by the end of 1947.
1948-50. “OF” is the only clue I have, so I have put Irvin in LF. He was playing corner OF and 1B in the majors in 1949-50, in any case.

Re Clarkson and Irvin in 1940:

From the Pittsburgh Courier 6/22/40: Washington, June 20-- . . . "Clarkson, new Eagle shortstop . . ."

From theChicago Defender 6/22/40: " . . . the club [Toledo Crawfords] is without the services of Bus Clarkson who has been loaned to Newark for the balance of the season. . ."

Thanks, djrelays!

Conveniently, a 6/22 start date for Clarkson with Newark fits a 2/5 SS, 3/5 CF division of Irvin's 1941 pretty well.

Monte Irvin estimated hit breakdowns

MLE
797 1B
200 2B
51 3B
140 HR

ML
504 1B
97 2B
31 3B
99 HR

Combined Total
1301 1B
297 2B
82 3B
239 HR

Hope that's everything Dan R needs to calculate WAR for Irvin!

Got any HBP, or should I add them in? And if I do, should it be in extra PA, or replacing a few current hits/walks/outs?

No HBP data for Irvin.

Extra PA should be added for the estimated HBP, I think.

Here are Chris's MLE's, translated into my WARP, filling in missing years using my war credit equations. Sorry they took me so long; between the multi-positionality and the mixed minor/major league seasons and the war credit, they took awhile to generate.



3-year peak: 24.0
7-year prime: 51.5
Career: 84.1
Salary: $249,366,905--close to Delahanty and Waner.

These seem stunningly favorable to Irvin. In particular, what stands out to me is that Chris has him as an average-plus fielder at his eclectic CF/SS/2B mix of positions over the course of the 1940's. If Irvin could really handle those difficult positions in MLB (particularly SS, where he's given 223 games) while posting OPS+ marks as high as 188, then he was almost an inner-circle great at his peak. Chris, how confident are you in this assessment of the quality of Irvin's Negro League fielding?

Chris, how confident are you in this assessment of the quality of Irvin's Negro League fielding?

I don't have much confidence at all in this assessment, but I don't think it any less likely than any other assessment. I had absolutely no statistical evidence to work with in making these projections: when it is a challenge to figure out what position Irvin was playing, there isn't going to be much in the way of fielding stats available . . .

All I had to work with were the facts that

1) it seemed to be the goal of Irvin's managers early in his career to get him out of the infield into center field
2) they kept playing him in the infield anyway, especially at shortstop
3) Irvin was clearly regarded as a great all-around athlete
4) Irvin was an above average corner outfielder in his thirties, even after significant ankle injuries
5) Former NeL players in the majors tended to be steered away from high-value defensive positions, even when it looks like they could play them well.
6) Good athletes who were contemporaries of Irvin were able to move between playing a decent third base and playing the outfield corners well -- exhibits being Mel Ott and Bob Elliott.

So I concluded that Irvin was surely not a whiz at third or short, but it was plausible that he would have been competent at those positions, especially during his early 20s. The one thing that I didn't examine, which I probably should have, was movement between third base and shortstop during this period -- I may have underestimated the degree of drop-off for Irvin in playing shorstop rather than third base. That's a statistical check that could be made to see if my projections are internally consistent with my assumptions.

But the point is that these estimates are based purely on what seems plausible, given a small body of anecdotal information.

If you want to see what a conservative projection of Irvin's major-league equivalent value would look like, I recommend doing MLEs using my most recent batting projections and the first set of fielding projections I did, which project Irvin to a corner outfielder career in the majors. That's the low-end possibility for his value, so that could show you how potentially optimistic these projections are.

That makes a substantial difference--as you know, in my system at least, SS replacement level is far lower than that of other positions. In the 1940's, we're talking 3 wins a year at SS, 2.2 at second, 1.6 at third, 1.0 in center, and 0.8 in the outfield (this was before CF really distinguishes itself as a mid-spectrum position in the 1950's). Clearly this is all guesswork, but I think my instinct would be to call him a -0.8 FWAA per year SS before the war, which would be the same total defensive value as an average second baseman, and a -1.3 FWAA per year SS after the war, which is a tiny bit better than an average third baseman. I'll also make him a league-average CF (rather than a somewhat above average one as you have him). So I'll peg his revised FWAA as the following: 0 in 1939, -.3 in 1940, -.4 in 1941, -.1 in 1942 and 1943, -.2 in 1944, -.4 in 1945 and 1946, and +.1 in 1947. That makes his total WARP for those years 1.5, 4.7, 7.1, 8.7, 7.0, 6.9, 6.2, 6.5, and 5.2, and puts his career WARP total at 80.5 and his salary at $233M, which is right next to Raines and Yastrzemski. Impressive! No change to his ballot placement for me--he's #6, ahead of Burkett.