I'm not going to argue it was the throw of the year. But in terms of timing and all that, it was spectacular.

While talking about that play I want to give some credit to Bobby Dalbec, who was playing 3rd, and came off the base to field the ball earlier so he could apply the tag earlier. I don't know if it made the difference but it seemed like a good heads-up thing to do.

Making the last out of the game at third is bad baserunning.

Renfroe has been a joy to watch out there. Santana....not so much.

The Red Sox are such a fascinating team this year, and this play is a microcosm of how unlikely the whole thing has been.

The closer in this critical game for Boston was...Hansel Robles, who has been pretty awful since coming over from the Twins. Even after last night's scoreless inning, he has an ERA of 6.14, and has allowed 28 baserunners in 14.2 innings. It is crazy that you'd use him in this game, with the playoffs sort of on the line.

He closed out the game started by Nate Eovaldi, who has been the best pitcher on the team this year. His problem was health, but he was absolute money last night, and is somehow currently second in the AL in innings pitched, with a 171/30 K/BB ratio. If we make the Wild Card game, he is honestly who I want starting that game.

The defense for the 9th involved a ton of positional switches. Bobby Dalbec was not the third baseman last night; Devers (as usual) was...but entering the bottom of the 9th, Cora moved everybody around. Here was the defense in the 8th inning:

C - Plawecki
1B - Dalbec
2B - Arauz
3B - Devers
SS - Iglesias
LF - Santana
CF - Hernandez
RF - Renfroe

Now, for the 9th inning:
C - Plawecki
1B - Shaw
2B - Devers
3B - Dalbec
SS - Hernandez
LF - Verdugo
CF - Santana
RF - Renfroe

The only two players who played the same position in both the 8th and 9th innings were the catcher and...Renfroe. (Devers at 2B - WTF?)

This is not a great team, but they absolutely nailed it on two low-risk, low-dollar free agent outfielders in Hernandez and Renfroe; have gotten exceedingly lucky with the health (if not the quality) of their top five starting pitchers; and whose bullpen has been saved by a Rule V pick from the Yankees who is probably (along with Eovaldi) the most valuable pitcher on the team this year. Devers and Bogaerts are elite young veterans, and they have been regularly playing at the top of their reasonable range of ability, keeping the wheels from falling off enough to be in the playoffs with 20 games to go. And they have not deviated an inch from seeing this as a season of rebuilding towards where they want to be in the next few years - a legit, sustained World Series contender. The farm system, ranked in the 20s last year, is now in the top 10, and is graduating two pitchers (Houck and Whitlock) who will likely be in the rotation in 2022.

This season is all gravy, from this fan's perspective. It has been one of the most enjoyable Red Sox teams to watch in my lifetime, and will leave me eager for 2022.

(Devers at 2B - WTF?)

Cora's stated rationale is that Devers has played on that side of the infield on some shifts, and has gotten used to playing there. The actual rationale is of course a combination of (a) desperation at 2B given most of their options had COVID and the one who has returned (Hernandez) is needed at SS, and (b) balancing time for Schwarber and Dalbec, both of whom have been hitting, and Shaw, who needs an opportunity to prove he can't. If Devers can cover 2B, Dalbec can play 3B and free up first for Schwarber/Shaw.

Making the last out of the game at third is bad baserunning.

Making the last out anywhere is bad. I'm not convinced this was bad baserunning. Having a runner on 3B with 2 outs is an advantage over having one on 2B. And it took a spectacular play to prevent that. Sometimes you take the chance and the other guy just beats you.

Arauz and Iglesias were pinch hit for by Verdugo and Schwarber. Could have kept Dalbec, Devers, and Santana where they started the game by putting Shaw at 2B (where he does have 41 games experience) and having Verdugo play CF. My guess is Shaw's gut rules out 2B and I guess Santana is better in CF? But probably not, and really needs to be released as soon as Duran is back from COVID.

Hansel Robles, who has been pretty awful since coming over from the Twins.

Yeah, cause he was so good for them too, this is really a shocker.

Making the last out anywhere is bad. I'm not convinced this was bad baserunning. Having a runner on 3B with 2 outs is an advantage over having one on 2B. And it took a spectacular play to prevent that. Sometimes you take the chance and the other guy just beats you.

The advantage is really small. And this was a nice throw from a guy who can chuck it, but not something otherwordly. It was bad baserunning.

Maybe the most amazing part of the throw was that Renfroe was running away from the infield as he picked up the ball, then had the wisdom to use a half second in loading up for the heave, instead of panicking and trying a quick "catcher throw".

The advantage is really small. And this was a nice throw, but not something otherwordly. It was bad baserunning.

Agreed the small advantage - 3rd vs 2nd with 2 outs gains little, as #8 noted. However, throwing a dime from almost 300' is something special. IMO, Meadows' baserunning on Monday was even worse. Despite the non-backups by the other outfielders, a decent throw/relay would've easily beaten him to the plate.

However, throwing a dime from almost 300' is something special.

For the average outfielder, it would have been an amazing throw. For an outfielder with Renfroe's arm, which is the only outfield arm that matters in this case, I don't think it was special. It had air under it and took a big hop to get there.

It's so weird that MS State only gave Renfroe 4IP to pitch in college, and he was awful even in that short opportunity.

Eh, it's a 35% increase in run expectancy (from .305 to .413). Not going to quibble about whether that's big or small.

It's not the Bo Jackson throw, or even the Cespedes throw. But it's a lot farther than I can throw a baseball.

Looks like a good throw to me. Not a great one. Virtually every Renfroe in the league could have made it.

Rick Ankiel, especially the 2nd one.

Run expectancy is the wrong way to address this in the last inning - you want win expectancy. I don't know the numbers for that offhand. I'd guess the break-even point is roughly a 90% success rate, with a rough guess that maybe 10% of the time the difference between 2B and 3B makes for a run (don't forget the wild pitch possibility.) It's quite possible that going was correct, that Renfroe doesn't make that throw 90% of the time, and this happened to be the 10% that he did.

That's a fair point but I'm not sure win expectancy is the right tool. They were trailing 2-1 in the 9th with 2 outs. Getting a single run keeps them alive.

Win expectancy is always the right tool within one game compared to run expectancy. The only reason to use run expectancy is as a simplified abstract tool, to be more tractable than calculating out all the possible continuations to get win expectancy.

That said, the mapping of run expectancy to win expectancy is very simple here, by using the distribution of run expectancy rather than the average. 0 runs = 0%, 1 run = ~50% (you could adjust with a generic or specific home-field advantage, or the talent of the pitchers and hitters expected to contest the extra innings), 2+ runs = 100%.

On the misplaced ball, which the runner saw, I'm siding more #6 than #8 on this one.

I don't know how people are parsing good throw vs. great throw, but I can tell you that watching this happen live without knowing the outcome, it was extremely shocking that he threw him out. He is picking up the ball close to warning track and the runner is already rounding second.

The play was set up by two misplays though:

--Santana appears to slow up then has to dive, and makes an ugly flop of a dive to let the ball go by him.
--Renfroe, who should be focused on backing up the play, is jogging toward Santana and has to reroute when the ball gets by him.


Win expectancy difference between runner on 2nd and runner on 3rd appears to be between 1 and 2 percent. You should be really confident you're going to make it to take that risk.

I was watching the game when this happened, and I honestly thought there was a better chance he'd go all around the bases when Santana did his crappy belly flop than there was that it was a two-base hit. I presume the baserunner got guidance from the third-base coach, and I presume the coach told him to go for it.

Anyway, the throw was spectacular, but what makes it really "wow" is the context.

Rick Ankiel, especially the 2nd one.

I kinda like the first one, because of Willy Taveras' reaction. Taveras had a +12 in baserunning that year, so that probably wasn't something he expected.

It's also the case though that, if it barely matters if the runner is on 2nd or 3rd, then we have to question Renfroe's decision to throw and risk the ball getting away, allowing the runner to score. He gets the greater benefit (game over) if he succeeds but it still may not be a wise decision.

B-R does give a second-by-second WPA but it was apparently at 96% when Wendle came to the plate. Throwing the ball away would bring that down to around 60% (somewhere around there I think) but obviously it's substantially below 96% at the point Wendle decides to try for third. I have no idea what the probability is of a bad throw (or missed catch or ball getting away while trying to catch and tag) but, technically speaking, you want to work those numbers before assessing Renfroe's decision. (My guess is the numbers work in his favor.)

Meanwhile, as I mention occasionally, on MLBtv in Australia, I get one of two gambling commercials and lots and lots and lots of the same 30 highlights repeated over and over and over during each commercial break. Based on this I can tell you there have been three miraculous throws over the last decade or so because I see each every game I've watched for the last few years. They are the Cespedes throw, the Laureano throw and the Robles throw (after an outstanding catch). I wouldn't exactly put any of them in the league with your standard outstanding Barfield or Dawson throw even if they were more "miraculous" in some sense. All three are desperation throws, 2 with no risk involved. All three throws have a good bit of air under it -- fair enough, they have a long way to go. Both the Laureano and Robles throws were to double off a runner who had rounded second -- miraculous but essentially "I've got nothing to lose by chucking it" and those just happened to be the 1 in 100 times the throw was on target and in time.

Obviously distance is a key component of a great throw but velocity and arc play into it and I'd be more impressed by, say, a laser throw from RCF that cuts down a Billy Hamilton going 1st to 3rd on a single than a desperate heave. The Renfroe throw here is somewhere between the two.

Does the win expectancy change materially because this is Fenway? In most ballparks, you can be pretty sure that Wendle is going to score from second on a base hit. I don't think that's true in Fenway where the left fielder is pulled in so far. So getting to third has a bit more relative value in this particular instance.

Right. Generally speaking, if an outfielder dives and misses the ball, the batter should be thinking 3 bases. There are exceptions, but not on that play.

It was Renfroe, and Renfroe is awesome at this very thing. But he had to make a perfect throw to have a chance. Too, the more I watch the clip the more I'm convinced that Wendle would have been safe if Dalbec had stayed at the base to wait for the throw. Basically Renfroe *and* Dalbec had to be perfect to get Wendle out. Renfroe being perfect wasn't enough.

Obviously the results don't bear it out, but I think it was worth the risk. I have a preference for aggressive baserunning, so YMMV.

There's no question it's a great throw. My point is, how great is it for Hunter Renfroe? It would be a miraculous throw for Johnny Damon. It doesn't strike me as too spectacular for him. He has a very strong arm, and it's pretty accurate (he's led his league in outfield assists two of the last three seasons).

But Walt's mention of the error is a good point. If this were Vlad or Dave Parker, who had cannon-like arms but were quite scattershot, then there's a better case for going for it because of the decent possibility the ball ends up in the dugout. Renfroe would seem somewhere between those two and the strong AND accurate arms of Barfield or Ichiro.

I have a preference for aggressive baserunning, so YMMV.

I'm the opposite, which is likely coloring my decision. I mean, I love it as a fan, but not from an analysis perspective. Of course, if this had been the second out, I'd have had no problem with it. I really don't like making the third out on the paths when you still have to rely on another offensive event to score (and, FWIW, though there is truth to what JAHV says about Fenway, I suspect the PB/WP possibility is less likely in this scenario, as the defense is acutely aware of such an event scoring the tying run and will be guarding against it more than they otherwise might. It would be interesting to see if that's true).

--Santana appears to slow up then has to dive

Santana got a poor read and slow jump on the ball to start. Any decent CF catches that standing up.

No one has yet mentioned that Renfroe nailed Margot at 2nd earlier in the game who was stretching a single into a double?

Quite the game for Hunter. As for "the throw", 300 feet on the money, one hop....c'mon, that's an awesome throw no matter who's throwing it or too much air, etc. It's a great throw.

It’s not a Rick Ankiel great but it’s pretty close

At Fenway, a single to left doesn't always score a runner from third.

EDIT: as Gary mentioned above.

Using empirical data; gaining 3b would increase the WExp by 1.3% Dying there, ending the game, the difference is 9.9 So you really need to make that base 8/9 times.

Empirically I used the last 30 years of MLB situations, but you could go back 60 or more and it wouldnt change much. THis site:

https://gregstoll.com/~gregstoll/baseball/stats.html#V.1.9.2.4.0.0.1989.2020

It often differs from BRef WExp, sometimes 5% or more. It uses real life home field advantage for instance which BRef doesnt.

Walt brings up a really good pt. about overthrows there in 24. I havent factored that in yet.

The diff in WE if the the runner scores is 28% or so. I think you'd still need a 33% error rate to make this worthwhile (I think??)

B-R does give a second-by-second WPA but it was apparently at 96% when Wendle came to the plate.

96.8% per the above site; the diff is probably the Home field advantage which is still about 2% in extras (HFA is 3.8 or 3.9% at game time)

--Renfroe, who should be focused on backing up the play, is jogging toward Santana and has to reroute when the ball gets by him.

why should Renfro be backing up? His first priority is to make the catch and end the game. Obviously Santana was not a gimme to make that catch. Obviously at some pt it becomes apparent that its Santana's ball to catch so only then does Renfro go into back up mode.

Eh, it's a 35% increase in run expectancy (from .305 to .413).

WHere are you getting that from? Run expectancy, the predicted number of runs that will occur from that base/out state, goes up somewhere between .035 and.04 depending on which data set you use. Per this site:

http://www.tangotiger.net/re24.html

Looking at Tango's tables there, I dont see 305 or 413 anywhere..

I suspect the PB/WP possibility is less likely in this scenario, as the defense is acutely aware of such an event scoring the tying run and will be guarding against it more than they otherwise might.

I respect your reasoning here and elsewhere but is that really a thing? Like you can't station the LF behind the C or something. is there any evidence that PB/WP occur less with tying run on third?

I got it from Fangraphs here.

I respect your reasoning here and elsewhere but is that really a thing? Like you can't station the LF behind the C or something. is there any evidence that PB/WP occur less with tying run on third?

That's why I asked. But catchers, like all players, undoubtedly lose focus over the course of nine innings. My guess is that if the tying/winning run is on third, they're going to be reminding themselves of the need to block anything in the dirt (and the pitcher of the need to avoid throwing one there), reducing the numbers of WP/PB that occur in those situations. But it's just a guess.

It was a great throw, no doubt, and a huge win, but it wasn't even the best throw to nail a Rays runner at 3rd in the past three years.

people can't just enjoy something. great throw, good tag, good baserunning, good baseball.

If this were Vlad or Dave Parker, who had cannon-like arms but were quite scattershot,

Farther back, this describes Rocky Colavito, maybe the most powerful arm I've seen. At one of his MiLB stops he'd do a pre-game exhibition of throwing the ball from behind home to beyond the CF wall 400' away. 300' throws on a line from deep right to home were always a possibility, but so were 350' throws into the box seats.

people can't just enjoy something. great throw, good tag, good baserunning, good baseball.

Why can't people just enjoy a good baseball conversation?

Seriously, it is possible to enjoy the play and still have opinions on the throw, the decision making, etc. What the hell are you here for if that's too much for you to handle.

Speaking of Hunter Renfroe, he seems to have created some controversy by saying that MLB told the Red Sox to stop testing for Covid:

Major League Baseball on Thursday denied Boston Red Sox outfielder Hunter Renfroe's allegation that the league told the team to stop testing for COVID-19 amid the outbreak spreading through its clubhouse.

"MLB basically told us to stop the testing and just treat the symptoms," Renfroe said earlier Thursday on WEEI's "Merloni and Fauria" show. "And we were like, 'No, we're going to figure out what's going on and keep trying to keep this thing under control.'"
. . .
The Red Sox refuted Renfroe's comments. "We have been following MLB's COVID-19 protocols all season long," the team said in a statement. "We have consulted closely with them on everything we've done and continue to test and their medical staff has been very supportive."

As of Thursday, nine people with the team have tested positive for the virus, with 11 others sidelined due to symptoms or close contact with positive cases. Among the players sitting out are Xander Bogaerts, Enrique Hernandez, Christian Arroyo, Matt Barnes and Nick Pivetta. The Red Sox are under the 85% vaccination threshold in their clubhouse, although some of the players who tested positive are vaccinated.

The article only mentions Renfroe, so one wonders how he came to his conclusion. Weird.

How can anyone tell if it was good/bad base running when all we can see are zoomed in views? A pulled back view of the play developing is vital. I need to see the ball, the fielder, and the base runner all in relation to each other at the same time. Not a video where the fielder takes up the whole screen, then the baserunner taking up the whole screen, etc etc.

I got it from Fangraphs here.

Hi Mefisto: It's hard to reconcile both tables. Most of the fangraphs RExpt are lower. BUt some of the cells are higher such as 1/3rd with two outs and 1/2 with one out. I dont think there's any way to reconcile by picking and choosing various time periods on the Tango matrix.

It seems the fangraphs one was built using a theoretical model based on a 4.15 runs/game environment. THe Tango one I guess is build on actual results. Very interesting.

It never even occurred to me that there would be much difference. I just grabbed the Fangraphs table and figured it was definitive. Very interesting indeed, that the differences are so stark.

How can anyone tell if it was good/bad base running when all we can see are zoomed in views?

It's a remark on the baserunning decision, not a remark on the technical aspect of the baserunning (and if the decision to go for three was based on the third base coach's recommendation, that would fall in there too).

Making the last out of the game at third is bad baserunning.

The only way he is out is by a near perfect throw and a quick ass smart tag, that play probably had a 95% success rate if not higher. That isn't bad baserunning, that is spectacular defense.

The only way he is out is by a near perfect throw and a quick ass smart tag, that play probably had a 95% success rate if not higher. That isn't bad baserunning, that is spectacular defense.

Bullshit. Hunter Renfroe has one of the best outfield arms (more than 30 assists in the last 300 games or so), and has made better throws than that. You have to consider the arm of the outfielder when you're deciding whether advancement is a good decision. This would ahve been a swell decision if Juan Pierre had gathered up the ball. Against a guy who has gunned down 30 guys over the last two-plus seasons, it's decidedly less so.

Super fun play and I'm glad everybody "went for it" even if both running (and possibly trying to gun him down) were potentially poor percentage throws. I like exciting baseball and that for sure was exciting. Plus the Red Sox won so it's much easier to enjoy than if I was a Rays fan.

Farther back, this describes Rocky Colavito, maybe the most powerful arm I've seen. At one of his MiLB stops he'd do a pre-game exhibition of throwing the ball from behind home to beyond the CF wall 400' away. 300' throws on a line from deep right to home were always a possibility, but so were 350' throws into the box seats.

Barfield was the best I've seen, no question. I'd imagine a lot of his assists (from what I remember) are of the kind where the runner goes, "OK, Barfield is fielding it, and you know he has a cannon, but this time I know I've got him beat! No way he gets me...ooops." Hunter's was of the same basic type-would basically take a perfect throw to get him, and he got him.

At age 23 Clemente had 22 assists. So the following year he was down to 10 and you think "well, the league sure learned its lesson". But no: he then had 19, 27 (!), 19, 11, 13, 16, 19. I guess all those runners kept thinking "no way he gets me *this* time".

CLemente 1958

22 assists,

what's league avg? Im guessing 7 (this article says Gardner saved 5 runs above avg in 2017 so...
https://sportsinfosolutionsblog.com/2018/06/11/why-does-brett-gardner-lead-outfielders-in-defensive-runs-saved/

So Im guessing 15 above avg VALUE 13.5 runs.

base runners held, 5 above mlb avg, VALUE 1 run

singles into doubles, last I seen it seems to track closely with BR hold, VALUE 1 run.

Range: 2.47/game vs league 2.02

2017 Betts had 2.42 vs league factor of 2.1; Statcast gave him 19 OAAs and 17 runs; OK lets give CLemente the same 17 runs

TOAL 32.5 runs

Total Zone: 27 runs

That's not too bad an estimate but it begins to show some of the limitations in TZ. If the all the difference came on range, they may have given Clemente 11.5 for range, where it was likely 17 or so.

Bullshit. Hunter Renfroe has one of the best outfield arms (more than 30 assists in the last 300 games or so), and has made better throws than that.

I think this is being a little harsh on CFB. Looking at the play from ground level, we would see the CF falling down and the RF running toward the wall to get the ball. Instinctively that tells you 3bases. And the coach had perhaps 1 sec. to process all that info and remember that Renfro has 16 assists or wotever this year.

You could put Clemente there or Barfield there and have them running back to get a ball and its really hard to figure how long they're going to pick the ball, stop they're momentum and then fire it to third. For instance how would we know how fast the ball is bouncing away? Do we know Renfro's exact distance? all that would have an impact on the timing in all this.

In retro spect and w/ cost benefit analysis its a mistake, in real time that's hard to figure.