But I was going to use Tommy John as my exception.
I actually agree with everything you say, I will almost always trade a low K pitcher for a High K of similar ages, but I also worry about factors like injury record and ptiching motions. Jarad Wright when he was with Cleveland looked like he would become a very good pitcher, but a lot of pitching Scouts worried about his arm motion. There are plenty of other examples including more recently Dontrelle Willis has had some question marks next to his name for pitching motion. His is not as violent as Wright’s was.
As ground ball pitchers go, Wang’s ratio is more exceptional than most most. If you pour through the stats, you probably will not find many with a nearly 3 to 1 ratio. Using ESPN’s sortable stats. They provide the G/F Ground ball/fly ball ratio GB divided by FB. Wang is in very good company with only Webb, Lowe and Aaron Cook ahead of him. This does not make him an elite pticher, but it does put him in company with good solid 2 or 3 guys. Webb is an Ace but he is the only one of the 4 that does have a high K rate. I think with Wang, it is worth doing a wait and see as currently he is among the elite GB pitchers for the past 2 year. He might be another exception to the rule.
Whoa, whoa, whoa. I’m not at all opposed to sabermetrics. I think that it’s wonderful. But when JP Riccardi, who learned Sabermetrics under Billy Beane in Oakland, signed Burnett, it was because of his ground ball-fly ball ratio and his strikeout rate. As a Jays fan, I’m certainly not gleeful about that signing now.
Honestly, I was pointing out a pitcher who didn’t perform as well as What Exit’s metric would have predicted, and I offered a possible explanation of why.
I have no idea how you got the idea that I’m opposed to sabermetrics in general, nor how you could possibly read a sarcastic tone into that post. I do think that they spend too much time interested in averages and don’t pay enough attention to variance. But like I said before, it’s not that I think that the stats lie.
You have no idea how I read that the saberheads failed to consider all the relevant information as blaming the saberheads? If all you were doing was criticizing Ricciardi for applying a sabermetric principle (that he may not have fully understood, and which contained no guarantees of future performance anyway) ill-advisedly, then I apologize. That is not how I took your comments.
I’m not quite sure what mean by “they spend too much time interested in averages and don’t pay enough attention to variance”–who, specifically, are you addressing? Where do they say that they do this? It seems as if you could be saying that the problem is that they like to deal in general tendencies rather than predicting outlying performances-- which is correct, but I don’t see why you’d be complaining about that. Sometime you get burnt? So do investors in the stock market. Doesn’t mean you should keep your money under the mattress.
The point that I’m try to make here is that players whose performance varies wildly(like Burnett) aren’t very useful. You can never depend on them to win a big game for you; they’re as likely to lose it for you as win it. Give me a pitcher who consistently gives you 7 strong innings rather than one who might give you a 9-inning shutout or might give you a 6 run inning. I think that this is a bit of a failing of sabermetrics. It’s not a massive failing, because they clearly have the right idea. I just think that they would be better served to consider the consistency of a player.
But Burnett is a career .500 pitcher. I mean, it has nothing to do with “Variance” - he’s an average pitcher who gets hurt a lot. His game-to-game performances don’t vary any more than you’d expect from a .500 pitcher. He’s going to lose about half his decisions and win about half of them. How would that be any better for the Blue Jays if he was more “consistent” but still lost half the time?
There’s no surprise at all that he isn’t winning 20 games. He’s done in 2006 exactly what he has always done; pitch so-so and get hurt. The signing of Burnett was as much out of a panicked decision that they had to sign a big name starter, any big name starter, as it was a logical decision based on stats; Ricciardi was suddenly given a big pile of extra money and he was in a rush to give it to someone, and they picked Burnett over Beckett for some reason. Sure, Burnett had a high K rate; he also has always walked a lot of guys and his ERAs were all accomplished in a good pitcher’s park. He’s 29 years old so there was no reason to expect a huge improvement.
Can you show evidence that some pitchers of essentially equivalent seasons or career statistical levels have different “Variance” in terms of their game-to-game performances? I’ve never heard of such a thing.
That reason being that Burnett was a free agent and wouldn’t have cost the top prospects that Florida wanted in trade for Beckett. Beckett has his own injury history as well.
Beckett’s hardly lighting up the league himself, his record aside.
Getting back to Burnett, the thing about him is that he’s either very bad or very good. He really is wildly inconsistent. So the theory that I was proposing is that you’ll see a high standard deviation in his sabermetrics. A quick look at his 2005 stats show a pretty high standard deviation in his K/9IP and his WHIP, but his GB/FB is pretty consistent.
Well with a quick lookup of Burnett on ESPN, he only had one year as a very good G/F pitcher. Lifetime he is only 1.46 but in 2005 he was 2.42. So perhaps Ricciardi was guilty of going by lasts years results or the hype that he was a good pitcher. I always thought he solidly overrated and as he was still a sub .500 pitcher, I would not have wanted to pay him. Of course I did not want Pavanno either.
Who I do want the Yankees to go after next year is Roy Oswalt. He puts up a large number of innings, gets his strikes outs, has a good injury record and low walks and home runs surrendered. The other non-statistical part is he appears to be a tough minded bull dog who goes after the win hard and should be mentally suited to NYC and the AL lineups.
Actually, scratch that; I don’t know what I was thinking. I guess I had it next to the K/9IP column, so 1.5 looked like it was a lot smaller than 4.3, but compared to his average GB/FB ratio of 2.4, the standard deviation was huge.
By correlation do you mean runs scored went down in the colder, northern climates, but held relatively constant in places like San Diego, LA, Arizona and Florida? Otherwise one could easily argue that the deviations in April and September were due to lack of timing and wearing down respectively.
Wouldn’t that just prove that it is harder to pitch in the AL. Granted it may be no harder to pitch and win, but one would likely give up more hits and runs. If this is true, Rysto’s point still stands. If an AL GM is trying to gauge how well an NL pitcher would do on his team he would have to take those added runs into account.
I’d also be interested to hear your definition of clutch. As far as I have always understood it, clutch refers to close and late situations, when the pressure is at its highest. I understand that runs early in the game are just as important, but, in my opinion, it’s asinine to compare at bats with two on and two out with no score in the second inning vs the same situation in the bottom of the ninth with the game on the line.
From what I saw on the ESPN stat site, there is no question that Ortiz has been a reproducibly clutch hitter for the past five years. In close and late situations his AVG and OBP look 30 points higher with RBI in 1/3 at bats (vs 1/4 regular season).
I think it’s hotter in Miami in August than it is in April. Obviously you have greater extremes in temperature in Toronto, but the principle still holds. If you want to do such a study, be my guest, but be aware that it could just as well prove my point as yours—and IMO much likelier.
No. That’s like saying it’s harder to pitch in Boston than in Yankee Stadium. It may seem harder, and pitchers have to work differently, but a .500 pitcher in Boston will not be a .600 pitcher in New York unless the Yankees have a much stronger lineup than the Sox do.
Oh–pitch and WIN. I’m starting to understand the goal of pitching now.
he would have to take a lot of things into account, but this one would reflect the differences in the situation, not the pitcher. Think of ERA as -/+ league average. That’s the figure that counts, not raw ERA.
A REPRODUCIBLE improvement in batting in situations where most batters do not improve. The capitalized word is the hard part for believers in clutch, because they like to point to a situation that has already occurred and infer “clutchness” from that. The problem is that in small samples (which can be surprisingly large by baseball standards) variations in results can seem significant without actually being so. Example. Let’s flip 100 coins. It’s not at all unlikely that in flipping 100 coins, I’ll get 53 heads and 47 tails. You might think this gives me special heads-flipping powers, but the real test of my ability would be if I can keep flipping heads beyond this point–at a point, IOW, when the results aren’t in yet. Theoretically, if I have clutch head-flipping ability, you shouldn’t have any problem betting that in the next 100 coin-flips, I’ll flip more heads than tails. But would you bet a lot of money on it?
In any given body of 600 coin-flippers, there will be one who shows the greatest results. Ortiz has been one of the most successful of the 600-odd MLB hitters over the last five years. He’s also a hell of a hitter in non-clutch situations, meaning he doesn’t have to show a lot of improvement to look pretty damned spectacular. The question is, if we can agree on a definition of clutch, will you bet that Ortiz will continue to show this improvement in those situations over the next x games? If yes, you’ve got a bet. if no, why not?
Key excerpt on clutch:
*Perhaps the biggest rap against Rodriguez is “he’s not a clutch player. He can’t win the big one.” Is there any evidence to support this?
Analysts have been arguing about the existence of clutch hitting for decades without being sure whether it exists or how exactly to define it. Bill James, the most influential baseball analyst, concluded after much study that “clutch hitting” as generally defined by fans and sportswriters is an illusion: Given enough chances, a player will hit in so-called clutch situations pretty much what he hits at other times. There are, though, different definitions of “clutch.”
Some people like to single out “late and close”: the seventh inning or later and three runs or closer. By this definition, there is a wide gap between Rodriguez and David Ortiz. From 2002 through 2005 in those situations, Rodriguez has hit .276 with an on-base average of .392 and a slugging average of .553, quite respectable numbers. Still, they’re not in the same league with Ortiz, who is .326, .408, and .724, though in fact Rodriguez’s clutch numbers are much better than the other Red Sox slugger he is often compared with, Manny Ramirez, who was .270, .387, and .423.
Many others look to postseason games as a yardstick for clutch performance, which is a handy stick with which to bop A-Rod, who hit just .133 in the five games of last year’s division series against the Anaheim Angels. Real fans, however, know that judging a player by a handful of postseason games is arbitrary and unfair. If Willie Mays’s reputation were dependent on his postseason play, it would come down several notches. In 25 games, Mays batted 89 times with one home run and a BA of .247 and an OBP of only .337. Was Mays not a “clutch” player? *
The part that measurement fails in fielding is range.Some guys just cover a hell of a lot more. Would they be more likely or less likely to make errors.
If I remember properly James had a dispute with bunting a man ti second. Giving up a certain out to advance a runner may not be statistically justifiable. But ,you will not likely get management to recognize it and change.
I think I read that a good clutch hitter hits the same with runners on base as with bases empty.
