Wednesday, May 22, 2013
Verducci: Offensive decline leads list of 10 early-season trends to watch
(sportsillustrated.cnn.com)
And not clicking on Verducci is quickly becoming another one!
Read More...1. Hitting in the major leagues is fundamentally broken
What will it take for teams to start admitting that this passive-aggressive, run-up-the-pitch-count philosophy isn’t working? Apparently almost a decade of declining results isn’t enough. Entering this week:
• The number of hits per game is down for the seventh straight year.
• On base percentage has been stagnant or down for the seventh straight year.
• Strikeouts ...
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
Page 2 of 2 pages
< 1 2The biggest thing that analytics has done in basketball is made much more clear what constitutes a good shot. In particular, the efficiency of corner threes and getting to the basket almost can't be overstated. Meanwhile, other shots that look pretty good to the naked eye, like 6 foot hook shots, are actually pretty bad percentage plays, and a defense should be happy if they can get their opponent to rely on them.
One thing that analytics have done for me is changed how I watch basketball. Even though I can't question the aesthetics of a pretty fallaway jumper from Kobe or Carmelo, I still cringe when they take those shots while there's time on the shot clock. Those are just plain bad shots, except at the very edges of game theory.
Yeah, I heard that recently in a Bill Simmons podcast with Zach Lowe (who is really good). He was thinking out loud that one of the reasons why Melo works better at the 4 is his being undersized for the position actually might incentivize that kind of shot. Pretty interesting.
The corner 3 thing is pretty interesting as well if only because now that everybody basically knows, I wonder if defenses will change to defend that possibility more and how that will in turn affect the corner 3 in the long run.
They already are...
How about some rebound study? Slowing up just a bit to let your offense get in position AND getting your players to crash the boards instead of giving up as I see sometimes.
Hmmm... I guess I can see how it reads that way. Let me clarify real quick - when I say that they come close to optimizing shot selection, I don't mean that they're as good at as they are with fully optimized coaching and strategy gamed out, I mean that they're probably only about 10% below optimal by just doing what comes naturally. The thing is, being 10% below optimal is enough to make you lose most games, it's the few points here and there that decide who's good and who's not.
I guess the problem is the imprecision of "comes close to". I mean that they're close in the sense that they're not taking awful shots very often, not that there's not big gains to be had in marginal wins by taking better shots.
Is this some of the Kirk Goldsberry work? I think he's shown pretty definitively that some "bad shots" aren't as bad as they seem, if there's rebounders in position.
There's a difference between "easy to model" and "big payoff." Baseball is easy to model for the reasons folks have mentioned (good stats, mostly batter-pitcher, independence of teammates) but it's difficult to model because the effect sizes are puny. A "20 point" difference in OBP is "huge" but is of course just 2% and is an extra base reached every other week. You need a couple thousand PA before you can start to detect that sort of effect reliably. (You can also think of that as the randomness of baseball outcomes.)
On the other hand, especially in the good ol' pre-parity days, despite football only playing 14-16 games per season, it wasn't that hard to predict which teams were going to do well. Or at least it didn't seem so to me.
Hockey is probably the hardest to model. Dependence, low-scoring and who's on the ice at any given moment is constantly changing (and while the puck is in play). Those stupid blue lines reducing the impact of speed. And my guess is the effect sizes are pretty small. But maybe it all comes down to the goaltender and whether you've got an offensive star.
OK, fast forward to the Billy Beane days in Oakland, and he's trying to identify talent that can help his team that isn't on the radar screen of teams like the Yankees so Oakland can get these players at an affordable price - and the main reason for the market inefficiency is because other teams were almost consciously not looking at the facets of play they should have been looking at (such as OBA).
Now, to my question - isn't there an absolutely HUGE opportunity in football to find information others don't have? Think about which offensive positions get emphasized - quarterback, running back, receiver - these are even referred to as the "skilled" positions, whereas apparently any no-talent schmo can play on the offensive line. Yet the battle at the line of scrimmage has a much larger impact on the outcome of games than anything else. If you have a great offensive line, literally anybody can be your quarterback, running back or receiver and you will put up huge numbers. For evidence of this you need look no further than the early fall portion of the college football season. Every year some Nearly Everyyear Highly Ranked University (NEHRU) gets a non-conference game against Podunk University (PU). Go to the sports books in Vegas and you'll see spreads of 50 points and above the mismatches are so great. Imagine pummeling a team 49-0 and you didn't even cover the spread! At any rate, you will often see second and third string players piling up the yardage for NEHRU just like the starters, because the offfensive line is just destroying the opposing team on every single play. With holes that big to work with, I could score a TD or two! But the press clippings - and the biggest paychecks when you get to the pro level - don't go to the linemen, mainly because we don't yet know of adequate ways to quantify their impact, leaving them largely undervalued. But I'm thinking some person way smarter than me is going to figure out how to do that at some point in the future and it will change the way the game is played and the way the players are paid.
Since Russell retired at the end of the 1969 season - has there been a single MVP in the league since then that was a "defensive specialist"? Not that there weren't great defensive players winning the award - Michael Jordan was a great defender, for example - but no one remembers Jordan for his defense, he is remembered for his scoring. I think every MVP since the days of Russell has been a great offensive player, either as a scorer like Jordan or passer like Magic or Nash.
Is the game such that nowadays no defensive player can have such a big impact as Russell did, even without being a great offensive performer? Is it the game that has changed that much? Or is it just that no one focuses on defense that much and that, if someone could properly quantify the impact, an inefficiency could be exploited?
This isn't generally the tenor of discourse with regard to NFL lineman. Left tackles are frequently very high salaried and very high picked. Here's some mock drafts for this year: http://www.cbssports.com/nfl/draft/mock
Notice that everyone's top ten is absolutely loaded with offensive linemen, defensive lineman, and linebackers that are pass rush specialists who will be essentially stand up linemen. There's a very talented corner, a couple reaches at QB (QBs really are so valuable that it's worth reaching), and that's about it in the top ten. A couple people thinking one receiver will go high. No one thinks any running backs are going high, and people have generally concluded that running backs were overvalued and are replaceable, with rare exceptions like CJ Spiller and Adrian Peterson. According to this SI piece, QBs are the highest salaried position, but second is defensive end, third is offensive lineman (not sorted by specific position), and fourth is defensive tackle.
So, yeah, maybe the casual fan isn't aware of the importance of trench play, but it's surely not undervalued by GMs. Also, there is some publicly consumable data that's emerging on line play, particularly at Advanced NFL Stats and Football Outsiders. This is in its infancy, but it's starting.
On Russell, I'm of the unpopular position that he's overrated, historically. He did a lot of things very well, was surely the best defender of his day, one of the one or two best rebounders, and a very good passer; the thing is, if you look at those Celtics teams, they were highly balanced teams that covered up for the fact that their best player was just an above average player offensively. I think it's hard to support the idea that Russell should have ever won an MVP, unless you buy the "best player on the best team" type of thinking. Sure, it's possible that he was otherworldly transcendent defensively in a way that one's been since, but that seems kind of unlikely, right? It's hard to really see him as being different from a better (offensively) version of Tyson Chandler or Ben Wallace.
That said, Tim Duncan's been the best player of last twenty years (if you don't count Lebron, since he has time left) and I think most people would agree that his defense and leadership stick out more than his offense.
Reaching for a QB is a waste. There isn't a top QB in the league that was a reach.
CJ Spiller on the other hand was a reach and a terrible pick by the Bills.
Assuming you are really being serious: Traditionally, UK footballers are awarded a cap, i.e. the kind you put on your head, for having made an appearance for their national team. Therefore, players who make an appearance for their national team, are known to have been "capped".
That's some serious hindsight analysis going on there. If they ended up being a "top QB", then pretty much by definition they weren't a reach. I bet there were about 31 teams who wish they had thrown a fifth rounder at Tom Brady though.
Thank you - I was not aware of that.
This depends how you define "top QB". Eli Manning, Phil Rivers, Joe Flacco, and Cam Newton were all reaches in the sense that they flew up draft boards late because teams needed QBs. On a draft grade basis, none of them were "worth" where they were picked, but they've all been good values in the long run. The very best QBs generally aren't reaches simply because their talent tends to be very obvious.
This can only be true in the sense that running backs are almost never good values. Spiller was the third most valuable back in the league last year and the most efficient on a per touch basis (http://www.footballoutsiders.com/stats/rb). While I'd personally never take a running back in the first round, that's an organizational philosophy thing, as I think they're fungible. Strictly on his own merits, Spiller's a truly special running back. The Bills miss a lot more than they hit, but they hit that one.
I was going to guess Kevin Keegan.
It's not hindsight. None of them were reaches at the time.
I don't consider Eli Manning, Joe Flacco or Cam Newton a top QB, though Newton could turn out to be. Phillip Rivers was until the past couple of years, he could rebound now that AJ Smith isn't there to do as much harm as good to the rest of the team.
The Bills had no passing game but they took a running back with a top ten pick, then to make matters worse they didn't play him his rookie season and as a backup in his second year. Two picks later was Anthony Davis, who would have helped the Bills more then and now.
I guess I'm not clear what the point is then; not every team can have a literal "top QB" then if you're defining it as something like the top 5 QBs. You need a QB, you need a pretty good one, and it usually takes a high first rounder to get one. Hell, if the Bills had acted on that basis in the past, we probably wouldn't have to rip them for the Spiller pick. Because of the premium placed on QBs, getting even a top 12 type guy usually requires spending pretty high picks.
The Bills are horrible organization in many ways, you'll get no argument from me there. I desired a small trade down and Mike Iupati in that draft, he was the guy I was most sure was going to be an impact lineman. The various ways that they Bills have been brutally stupid with their selections, signings, and usage patterns doesn't really have any relevance on evaluating Spiller's abilities though. He was an awful pick because of the context, I hated it more than just about anyone, but he really is a special running back.
I disagree, but I'm a bit biased. I really liked him in college and then I've been able to watch him develop. But I also wouldn't call him an elite QB.
That when you reach for a QB in the first round "because it's a QB" you wind up with crappy QBs like Christian Ponder, Jake Locker, Brandon Weeden and Tim Tebow, which only does your team more harm than good because you don't improve and it costs you your job. Of course, not every QB that is drafted high and is universally agreed upon to be a worth the pick turns out to be any good, but the guys in the league who were first round reaches all stink.
Well yeah, that's what makes a player tops at his position. If I wanted to say starting QB I would have said starting QB.
But, yeah, once Oscar was in the league I'd have a hard time believing Russell was really the most valuable or productive.
By the way, for your enjoyment/disgust, the bball-ref ELO ratings:
#27 Kevin Durant
#28 Scottie Pippen
#29 Oscar Robertson
#30 Bill Russell
#31 Dirk Nowitzki
#32 Wilt Chamberlain
#33 Jason Kidd
Edit: Wow! Even more bizarre than I thought ... David Robinson at #3 all-time??
Edit 2: And this just in -- Dominique Wilkins just passed the Big O. That's the first pass of 'Nique's life.
Boxers and trainers already glean that sort of information from watching video. Adding any sort of quantification to what they intuitively pick up as part of pre-fight preparation wouldn't be of much, if any value in my estimation. If I know a fighter likes to double-up on his left hook, body-to-head, the number being 50% of the time or 70% of the time doesn't make a difference, since I'd be training to counter it regardless - these tendencies become part of a fighter's established style. Boxers have been dissecting film to pick up on subtle tendencies and tells for 100 years now, I don't think adding any sort of reductive analytical aspect to the study adds any real value.
There may be some potential to a deeper study of the biomechanics of the punch, but even that has largely been broken down effectively over the long history of boxing. Jack Dempsey wrote a book, "Championship Fighting", that is absolutely magnificent in effectively describing the biomechanics of proper punching technique.
Now if by "Moneyball" you mean, instead of a quantitative approach to analyzing talent, ways to exploit inefficiencies in the overall market, you could have made a mint just betting on the guys Frankie Carbo was putting money on in the 1950s.
'cuz white guys post up like THIS, but black guys post up like THIS.
I too saw only the last couple of years of Russell's career, and the centers I remember best from those years were Chamberlain and Reed, who were pretty damn quick. But they were not typical, either. I looked back through NBA-Ref, and the more typical center of the day was Darrall Imhoff, who I remember as a great lumbering individual - and he was a *good* player, a great big pillar of a center. Indeed, Russell matched up very well against such players.
But they absolutely have stats on how each lane plays during a tournament.
When Morgan was younger he was graded a C class player. Didn't get graded outstanding in any aspect of the game (and the scouts managed to avoid noticing that he was always on base. 148 walks in 148 games and no mention.) Didn't get graded as A for speed either and frankly, missing that means that the scouts in question simply weren't paying attention.
As late as 1970 the Senators scouting reports were dismissive of Morgan. Swear to God they were saying that they didn't need him. This from the team whose 2B job was in the hands of Tim Cullen and/or Bernie Allen (Not that the two combined wouldn't have been a fine player. Tim Cullen was a fine defensive player who couldn't hit despite a great deal of work with Ted Williams. Allen wasn't a bad hitter for a middle infielder of the day)
Hell, Morgan was a NP (non prospect) when signed. Second player ever signed by Bill Wight (The first being Walt "No Neck"
Williams 5' 6", 180) You know how the scouting community reacted? "First Williams and now a midget." They're laughing at
a guy whose first two signings make the majors and one of whom is an inner circle HOFer.)
There was a Ranger scout who wrote in 1977 (ie after two consecutive MVPs) that he was "Somewhat limited because of small body."
The problem being that a player has somewhere around 200 total shots by the end of the round robin so a the third on a team that hits a lot might have 3 in turn draws all week, And if somebody blows a sweeping call, or one of those shots picks ...
While ice conditions at the top events are pretty consistent these days (basically only two people do the ice at major events) so you could get meaningful numbers for any of the top teams.
I bowled in a competitive league. I was one of the poorer bowlers, averaging in the mid-180's. I bowled in the county tournament twice, which was prepared under professional conditions. They had a name for it, I don't remember what it was. Anyway, I struggled to crack 120.
The Elo ratings at the top pretty fluid, and can change drastically at just the whims of a few voters -- I'm not sure how much weight I'd put into them.
(I love this site -- I get the opportunity to learn so many random things)
Here is a list of the worst contracts in the NBA according to Simmons. So you would think that "Moneyball" might work for the NBA but it seems that there is always going to be a NBA GM that will overpay a marginal player.
I believe it's how they oil the lanes. Lanes for recreational bowlers have more oil in middle of the lane, where you want to be, so the ball will roll truer. I read something where this has changed over time (bowling alleys are intentionally making conditions easier to attract more bowlers) and 300 games are now wildly more common than they were 30 years ago.
snapper's right, it's how they oil the lanes. If you throw the ball without any sort of hook at all, it won't matter. But you also won't get many strikes that way. What most people don't realize is that bowling lanes are very, very oily. If you threw a hook on a lane with literally no oil, the ball would go shooting off to the left (for a righty) within about 2 feet from the line. There's a lot of spin on a ball, even with only a small hook.
So for recreational bowling, they oil the lanes heavily in the middle and drier on the outside. If you miss your spot to the middle, it hits more oil, doesn't hook as much, and you still hit the headpin. If you miss your spot wide, it's drier, and your ball will hook more, and you still hit the headpin. In a recreational league, you basically get huge error bars. In a tournament, they spread the oil more evenly, or even the opposite (drier in the middle and more oily on the outside), and you have to hit your spot exactly to get the strike. Miss by even a board and you miss the headpin altogether.
There's also a spread of how far down the oil goes. Sometimes they let it be dry close to the pins and you'll see balls go down without much movement, then take a huge left turn near the pins. Again, it makes it so that you have to throw exactly right to get the strike.
And yes, there are analytics for this stuff.
I actually had a lengthy conversation on this topic with a couple of very well-schooled boxing historians several years ago, and I do think that there's an answer to your question and that this answer has changed significantly over the decades as the economics of boxing have changed.
Yes, true, styles make fights, and by the 1930s at the latest boxing technique had settled in around the tactics and limitations created by the universal adoption of "small gloves", enough so that a handful of consistent "styles" had become fairly codified - your "sluggers", your "swarmers", your "technical boxers", and your "boxer-punchers". To some extent there's a bit of "rock-paper-scissors" in matching up the styles but every fighter, style aside, is an individual who his own strengths and weaknesses, so while it is generally fair to say "Technical boxers beat sluggers (Ali d. Foreman), swarmers beat technical boxers (Frazier d. Ali), sluggers beat swarmers (Foreman d. Frazier)" it's really only a generality akin to a baseball addage like "left-handed sluggers like the ball low and in."
But the really interesting question is how you develop your young, greenhorn fighter to cope with the varying techniques and styles he'd encounter as he climbs the ranks to eventually make real money (for you, the kindly manager, of course). In this aspect there has been a series of enormous changes in the convention wisdom over the last century. Working backwards, for the last 30 years or so as PPV has come to entirely dominate the marketplace of the sport, the most valuable thing a young fighter can bring to the bargaining table is a perfect record. That gaudy 25-0 allows a promoter to crow to the heavens, "UNDEFEATED!" as he'll try and invoke the memories of a young Tyson, a young Ali, a young Hagler, or any number of all-timers who won belts without a blemish on their records. The fighter who sports a mere 20-5 record against superior competition, including dipping his toe into the top-10 waters for a couple of losses, just doesn't get that kind of scratch - worse, if he's actually learned a thing or two from his experience, the 25-0 fighter's manager will try to keep away from him, 'lest he screw up his golden boy's gilded record and the casual public's dream of what he may become.
I think it's obvious to say that fighters simply didn't have that sort of luxury in the "old days", when the field was deeper as was the public's interest in the sport. Up until the advent of televised fights the preferred method of developing a fighter was to have him fight as often as physically able against a variety of styles specifically for the purposes of putting a few beatings on the lad to "get him serious". Many fighters simply took as many fights as they could schedule, opponent be damned, because they simply needed the payday - I consider Jersey Joe Walcott their patron saint. This man became heavyweight champion of the world at 37 (and was robbed of a deserved championship win over Joe Louis some years earlier). Look at his record. Look at the records of his opponents. This was a blue-collar championship rise through the ranks, against hungry fighters eager to play spoiler for the promise of an extra $20 on their following fight. If you really want to bug your eyes out, look at where the name "Elmer Ray" shows up on Jersey Joe's record, and imagine what Elmer's career was like (he was a scary dude).
Anyways, I guess my point is that the "ideal risk/reward" for bringing up fighters has changed a fair bit. Fully admitting I don't follow boxing anymore I don't see any reason to doubt that it's still focused on protecting undefeated records for rising stars in hopes of getting at least one glorious payday out of them. Just glancing at the latest series of victims of heavyweight champ Vitali Klitschko I see his last 4 defenses were against fellows with records of 21-0, 15-2, 44-1, and 17-0 - 3 protected greenhorns and a blown-up lightheavyweight, basically. Joe Louis' last championship defenses were against Jersey Joe Walcott and Tami Mauriello (69-7-1), Billy Conn (62-10-1), and Abe Simon (36-9-1)*. Pretty big difference in experience there. Promoters don't get giddy at the thought of enticing casual fans to buy a PPV to watch a title defense against a guy who has 10 or so losses already.
* If you BoxRec Joe's defenses you'll see the name "Johnny Davis" in that string, which I omit. The fight was scheduled as a 4 round exhibition, and at the very last minute the NY State Athletic Commission demanded it be changed to a legitimate bout, claiming they did not recognize "exhibitions" in their jurisdiction. Davis was a very low-level fighter and would never have found himself in the ring with a fighter of Louis' caliber were it not supposed to be an exhibition, as the newspaper report of the fight bears out.
Majority of plays in Football have a defined start and end, with an outcome that can measured in yardage if not simple success/failure. There's also the nice relationship that positive yardage fairly directly contributes to scoring. In Soccer and Hockey, you have the added difficultly of the play being fluid with few defined stoppages (and in Hockey you even have player substitutions while the play is still going on). Plus you have the relatively huge question of what exactly goes into creating scoring opportunities.
I was looking at Harry Greb's record the other day, and I was just totally blown away by his fight schedule. He was fighting top guys, fighting up one or even two weight classes, and taking new fights every two or three weeks! And it's not like he was getting out of those fights unscathed, either...
There are some broadcasts out there dating back to 1963, all Finals games. Russell runs so relatively fast and jumps so relatively high and his timing and sense are so great that it looks like he's playing a different game than the other players, very much like Orr or early Gretzky.
At least one of the games is against Wilt (SF Warriors), but Wilt is so lumbering and "low-posty" that the point still holds. In terms of fluidity and speed and grace, he wasn't in Russell's class. (Small sample size noted.) Other than not stepping out to three-point land, Russell's game looked thoroughly modern, even though it's now 50 years ago.
Oh yeah, even by the standards of his day Greb was considered a freak of activity (more than 40 fights in 1919 alone!). Here's his official record on BoxRec, which you've probably already seen - 299 bouts in all. Some other famed iron men of the ring from that era were former lightheavyweight champions "Slapsie" Maxie Rosenblum and Battling Levinsky, whose prolific bout output rivaled Greb's, but Greb fought well outside of his weight class routinely and at the highest levels of compeition - it's no knock against either of these all-time greats to say they don't quite measure up to Harry Greb and they'd probably say the same thing - Harry beat 'em both! Greb had no "decline phase" either, dying during minor surgery at the age of 32 with plenty left in his tank.
The real question here is this - was Harry Greb the greatest pound-for-pound fighter to ever lace up the gloves? Does his career overshadow those of ring legends like Sugar Ray Robinson, Henry Armstrong, Willie Pep, and Muhammad Ali? The record, in my opinion says "yes", but my answer to the question is a resigned "no". Sadly, in one of the great ironies of the sport, every single recorded frame of Greb in combat has been lost to time. The man with 300 career fights left not a second of footage for posterity, and the man who hated to train can only be found on film in a series of training clips taken for newsreels in the leadup to his legendary 1925 fight with legendary welterweight champion Mickey Walker (view every surviving second HERE along with a wonderful write-up from the great Mike Casey). The man who defeated dozens of all-time elite fighters, and whose whirlwind style was considered absolutely singular and unlike anything else the sport had ever seen, has left us nothing for analysis, and for that I cannot in good conscience rank him where he probably deserves to be ranked as an all-time fighter.
Page 2 of 2 pages
< 1 2You must be Registered and Logged In to post comments.