Filter posts by...
Leverage_Index
Monday, August 11, 2008
Your reliever gets an out. Your chances of winning go up. Your reliever allows a runner on. Your chances of winning go down. You add up the deltas of all the times that your chances went up, and you add up all the deltas of all the times your chances went down. Call the former “Win Advancement” (WA) and call the latter “Loss Advancement” (LA). WA+LA is GA (Game Advancement).
If you start the game at zero, you are marching toward 1 Win or 1 Loss. In a typical win, the pitching team will accumulate 1.8 WA and 0.8 LA. The difference in WA and LA, for every win, is always 1.0. Always. That is, on your march toward a win, you’ll accumulate some good things and some bad things. And in a win, you’ll accumulate alot more good things than bad things. The difference, in a win, will always be +1. Similarly, in a typical loss, the pitching team will accumulate 0.8 WA and 1.8 LA, with the difference always being -1.
So, in an average game, you have 1.3 WA, 1.3 LA, 0.5 wins, and 0.5 losses. The WA and LA capture the ebb and flow of the game, on your march toward the win or loss of the game. There is, on average, some 0.8 “wasted” WA and 0.8 “wasted” LA per game (2.6 GA minus 1 game). In order to align WA and LA to W and L, simply subtract the waste (average of 0.8 wasted advancements on each side) from the total accumulation in each game (average of 2.6 GA) from each of WA and LA.
Before we talk about relievers, let’s look at the last generation’s four greatest starters:
Read More
Wednesday, August 06, 2008
This article by Caple has it all. A couple of wrinkles:
I asked Red Sox closer Jonathan Papelbon whether the definition for save situations could be improved, and he said no. “A save is what it is. You save the game. It’s a situation in which the tying run is at the plate or on deck and the game is on the line.”
If only that were true, we wouldn’t have much of a problem here. Caple doesn’t even correct Papelbon. Entering the 9th inning with a 3-run lead and bases empty means that the tying run is not at the plate and is not even on deck. He’s still on the bench. And I think at one point the save rule was limited to the “on base, at bat, on deck”. It’s all the darn exceptions that kill the save rule.
The answer is yes, of course, he can, especially if roughly two-thirds of his save opportunities continue to come with a two-run or more lead. K-Rod has yet to appear before the ninth inning.
He’s talking about KRod breaking the saves record. K-Rod’s “gmLI” (LI when entering the game) is 2.15, which is on the high end for relievers. His “inLI” (LI when entering each inning of the game) is 2.20, which is also on the high end. His “pLI” (LI at the start of each PA) is 2.55, which is very very high. The third number is more influenced by himself, in that he probably put himself in a tough position by walking batters then getting himself out of a jam. This seems to be his career pattern by the way. His LOB rate is very high (82.4% career, 83.0% this year). Even Mariano Rivera has a career of only 79.4%.
I don’t even have to check the split stats, as I can only imagine how KRod pitches: bad with bases empty, comes in with high LI, puts runners on base, gets an even higher LI, then punches out the side. Here, I’ll check anyway… gimme a sec....
.191 .273 .324
.177 .263 .263 (after removing IBB)
Yowza, what a high-wire act. Maybe he shouldn’t be pitching from the stretch. And he definitely shouldn’t sign with the Mets. Their fans hate closers like that.
Anyway, back to Caple’s main concern: since KRod’s gmLI and inLI are above average for a closer, he was not the best case to use. Billy Wagner and Papelbon are probably the worst-used of the closers. Though Wagner has so many blown saves (7 already), it’s hard to say that the Mets were wrong in his case. Hard to have as many BS as that, if he doesn’t come in much in high LI situations.
Wednesday, July 23, 2008
Bill James posted Mariano’s seasonal Win Shares - Loss Shares, whereby he gives him a total of 136 Win Shares and 24 Loss Shares. I replied:
Read More
Thursday, July 10, 2008
What WPA represents is the quantification of your feelings as the game unfolds. Imagine if Youk hit into a triple play his first 2 AB, with the score tied 0-0 the whole time, then the Sox lead 15-0 (and he gets two outs), then he hits two HR. How is it that you felt with Youk, if you tracked it in real time?
Well, his first two AB, you are cursing his name like there’s no tomorrow, then when the team batted around (twice), your blood pressure starts to go down, and then, with the score at 15-0, you’re probably not even watching the game any more.
That is what WPA captures.... the quantification of your feelings as the game unfolds, assigned to the players involved.
WPA is not a way to evaluate the talent of a player. WPA is exactly the same as counting a PA 11 times when the bases loaded down by 1 with 2 outs in the bottom of the 9th and counting a PA as almost zero in a blowout game. It is basically ridiculous to think that one PA can inform you on the talent level of a player 1000 times more in one situation than another.
WPA/LI however might be a way to evaluate the talent of a player, since now each PA is exactly worth 1 PA. The only thing we are doing is realizing that baseball might not be a random game, and that a player might tailor his approach based on the base/out inning/score. We don’t know how much he does tailor his approach. That needs to be studied.
Saturday, June 14, 2008
By , 07:34 PM
Mark Shapiro, the GM of the Indians, was chatting in the booth today during the CLE/SD game. He seems like a real smart guy. The Indians are a very good organization, so that is no surprise. He was talking about bullpens, and the TV announcers asked him why bullpens were so unpredictable. His answer was two-fold. One, he said, that there might be a bounceback effect, whereby relievers who pitch well are overworked and may be less effective the next year, and vice versa for relievers who pitch poorly. I did a study on that a while ago, which suggested that that is not the case, although, to be honest, I don’t think my conclusions were definitive. Shapiro did not seem to sure about that either.
The second reason he opined, was right on the money. He said that because relievers only pitch 70 IP at the most per year, that teams make a lot of mistakes in terms of evaluating them. While he was on the right track, he also should have mentioned that when we compare bullpen performance from one year to the other, we are comparing two sets of performances both of which have a lot of noise. By definition, that would create lots of unpredictability.
I have mentioned this concept before, and I think it is an important one: There is a big difference between 550 IP (about how many IP relievers pitch per team per season) thrown by 10 or 15 guys (the number of relievers per team) and 550 IP thrown by one pitcher, in terms of the variance of performance. A huge difference actually. When people think of variance, they tend to focus on the total sample size rather than the sample sizes of each player, where the performance of each of those players is pretty much independent of one another.
Anyway, I am going to answer the question of, “How do you build a bullpen?” which seems to be such a difficult and illusive answer, at least according to baseball insiders. It is not. And of course, I am not really sure what they mean by “build a bullpen.”
1) Find halfway decent starting pitchers that you think are also suited or even better suited for relief, and turn them into relievers.
2) Stop rotating relievers in and out of the major leagues, based on short-term performance. Keep evaluating your reliever projections, including your pitchers in the minors, using sabermetric techniques, and your scouting reports, and use them to determine who stays and who goes.
3) Sort your relievers according to their projections, and make sure that your better relievers get more high leverage situations and that your worse relievers get fewer high leverage and more low leverage situations.
4) Resist the temptation to use your better relievers when you are losing by 4 or 5 runs (or more), on order to “stay in the game” and resist the temptation to use your better relievers when you are up by 4 or 5 runs (or more) for fear of losing the game. Use your worse relievers in low leverage situations, period.
5) Use you righties and lefties wisely. In fact, each reliever should NOT have just one projection. They should have one projection versus RHB and another versus LHB. Use those to determine who comes in when. Also, rather than having just an “ERA” type projection for each pitcher, classify pitchers according to high or low K, BB, HR, and GDP (basically G/F ratio). Use those also to determine who comes in when, depending on the game situation.
6) Of course, us your ace in more IP per season, and use him in multiple innings when you can, and use him any time the leverage is high, regardless of whether it is a save situation or not.
7) Bring in relievers as early and as often as possible, especially when your non-ace starter is on the mound.
That is how you “build” (and use) a bullpen. It is not that hard.
Wednesday, May 21, 2008
Fangraphs is showing the average LI and average win expectancy for each game for a particular day in its sidebar (sorted for your ease). So, you can see which game is the most “exciting” based on these two measures (high LI means alot of swing potential, and a win expectancy that is close to .500 probably means that the game was close for a good part of the game).
Tuesday, May 13, 2008
You guys know I love the Wisdom of the Crowds approach. So, let’s use it for something that has an outcome, rather than just the academic exercise we’ve been doing. I offered for Leverage Index to be part of the Bill James Handbook.
For a non-commercial product (specifically, if one of my readers can get the product or service for free), I have a free licence for it. I figure that as long as you guys don’t pay for it, I’m happy to donate it to these guys so that you in turn can see it in action. That’s why I’m happy that Fangraphs and B-r.com has taken it.
Now, other commercial ventures have approached me for it (video games mostly). I’ve turned them all down, because now that’s a consumer-paid product (as opposed to the ad-driven service that Fangraphs and B-r.com offer), and I honestly don’t know what it’s worth. Plus, I never play video games, so I don’t really know how it’s going to be used, etc.
So, this is where you guys come in. What is a licence to Leverage Index worth? You can state it in terms of dollars, in terms of % of revenue, in terms of value per book sold. You can state it in terms of in-kind, like data(*). Or in other terms that you can think of. Maybe make Bill James agree to drop Runs Created in favor of BaseRuns! Whatever.
(*) For data, I already have an understanding that any work I do with BIS-related data at Fangraphs is published at Fangraphs. So, I’m thinking there is a limit to the value over and above this. But, maybe you guys have other ideas here.
Make me a deal guys. You are the wise ones.
Friday, May 09, 2008
By , 03:07 AM
Wed. night in the 9th inning of the Tex/Sea game (not that anyone would be watching that game), C.J. Wilson, the Rangers’ regular closer, came in in the 9th inning to preserve a 2-run lead. (Padilla had pitched a very good game, throwing 96 and 97 mph gas!)
Wilson pitches to the first batter and goes 1-0. Remember that Texas is leading 2-0, there are no outs and no one is on base.
Read More
Wednesday, April 16, 2008
I put this csv file out there several years ago. It covered the 1974-1990 Retrosheet years. I’ll be working on my Retrosheet database in the coming weeks, and anticipate putting out career totals for LI, though I suspect that baseball-reference.com will do that eventually. Until then, you can enjoy this blast from the past.
Note: the list includes starters.
Thursday, April 10, 2008
Geoff gives us the lowdown on one of our generations best relievers. I responded at ballhype:
B-r.com now has splits by save and non-save situations:
http://www.baseball-reference.com/pi/psplit.cgi?n1=hoffmtr01&year=0&lg=&team=#outco-outco
Three things jumped out:
1. lots of IBB in non-save situations
2. much higher BABIP in non-save situations
3. alot more multi-inning games in non-save situations
Wednesday, April 09, 2008
I furnished Sean my Win Expectancy and Leverage Index charts on the expressed condition that anything he publishes from it be given to the public for free. And here it is. Tim Raines, as we can see, is a Clutch Hitter. In high leverage situations, he’s above his career norms, despite the fact that he faces tough relief pitchers (who, by the way, he also eats for lunch… check out his splits against relievers). Sean’s got alot more splits too, enough to waste the rest of the week.
Mariano Rivera doesn’t do as well in high-leverage as he does in low-leverage situations. Note: watch those IBB. Goose, on the other hand, shows a clear split the other way.
Thursday, January 10, 2008
Jack Morris thinks that if he pitches a complete game in a win (presumably by a margin of score of 3 runs or less) that he deserves a save. I’m ok with that I guess. Morris has 111 complete games among his 254 wins. I don’t know how many would have qualified for a save, but let’s say that number is 90 or so. So, he gets 254 wins and 90 saves. Bert Blyleven is at 287 wins and 167 complete games, so let’s give him 130 saves. Smoltz has 207 wins, 43 CG in those wins (let’s say 30 “saves") plus 154 actual saves. From this standpoint, Smoltz doesn’t look so good. He’s got 94 total “saves” on Morris, but 47 wins short. Smoltz though, is a far better pitcher than Morris. Plus, he actually has had fantastic post-season numbers.
Dave Smith of Retrosheet posted some interesting data, which I simply aggregated as:
Read More
Friday, October 26, 2007
I was hoping somebody would do this. The sum of each event’s WPA/LI tells you exactly the impact of an event, without the effect of the leverage. If someone can move someone over, WPA/LI will capture it. And thankfully, Pizza was the one who took the plunge. Next time someone talks about “productive outs”, Pizza has the list as to who can do it. Hopefully, he can share a more extended list, with numbers.
Wednesday, October 17, 2007
Josh Kalk checks in again. I’ll have to think about the “end of inning” scenario he brings up.
Friday, September 21, 2007
Joe Sheehan’s turn to talk about Ned Yost, and echoes what has already been said. (If you want to talk about Ned Yost, go to that thread.) But, that’s not why I’m here. He talks about the “Leverage score” (LEV), and refers to this for Brewers. The correct Leverage Index (LI) are here. LEV is not LI, and no one should confuse the two, nor think that each has some advantage over the other. They don’t. LEV should not be used to discuss leverage of situations as Joe Sheehan is doing. Ever. This was discussed in…
Read More
Wednesday, July 25, 2007
Let’s see:
Read More
Friday, July 13, 2007
This from Joe P. Sheehan is pretty cool too.
Friday, April 27, 2007
There’s alot of good info in The Book on this. Here are a few more considerations that researchers can try to analyze:
Read More
Thursday, April 19, 2007
In the greatest feature column in America, a new thing we learn is that the Mariners have not been involved in any close game, according to Studes’ definition. This can be better expressed with Leverage Index. The average LI is 1.00. The Mariner hitters’ average LI is 0.71, their starters are 0.83, and their relievers are 0.31. Heck, their closer has an LI of 0.18!
All this of course is small sample sizitis. Eventually, the Mariners will be involved in some thrillers. The Yankees for example are playing in thrilling baseball games, with an LI of around 1.2
from Amazon
Recent comments
Older comments
Page 1 of 65 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date