Filter posts by...
Talent_Distribution
Tuesday, September 13, 2011
By , 03:22 AM
I know there has been similar research published on the web, but I am too lazy to look it up right now. I took the 1998-2010 draft list from BA and looked at how the various pitchers did in the major leagues, breaking the rounds down into various buckets. It was a quick study and I just matched names from the draft lists with my major league databases. I probably missed 3-5% of the players because the names, especially first names, did not match up exactly.
First I looked at the rookie years for all pitchers in each draft round 1-3+, for a total of 4 buckets. Each round included the supplemental rounds, so, the first round actually has 60 picks in most years. The rest of the rounds pretty much have 30 picks each. Perhaps I should consider the first 30 picks of the 1st round as the 1st round and then the next 30 picks in the 1st round supplement as the 2nd round, etc.
Does anyone know if there is anything special about the supplemental round after the first 30 picks or is it just the next 30 best players and then the second round is 61-90 best players?
For each pitcher, I looked at whether their primary role in their rookie major league season was as a starter (S), a reliever (R), or mixed (N).
The currency I used was my normalized, component ERA (nERC), which is an “ERA” based on a pitcher’s raw stats (s, d, t, hr, nibb, hp, outs, and wp) adjusted for park, opponent, and defense, and normalized to his own league, where the average pitcher in each league, weighted by TBF, is 4.00. I weighted the aggregate nERC in each bucket by each pitcher’s IP. If I used the simple average of all the pitchers (weighting each pitcher exactly the same regardless of how many IP they threw), the numbers would be much higher, as the pitchers who had the worst true talent generally had the fewest IP.
Read More
Friday, September 09, 2011
The primary reason something stabilizes is based on the spread of true rate of whatever entities are involved.
As an example, hitter K/PA stabilizes very fast because there is a wide range in talent in terms of striking out.
The reason you find the wide range in talent is if a player is selected based on that component.
For a nonpitcher, having a high or low K/PA is not that important, because there’s tons other things a nonpitcher brings to the table to counteract the negative effects of a high K/PA. So, by not selecting specifically on that, we see a wide range in talent.
For a pitcher, having a high or low K/PA is important, because there’s comparatively little else he can do. And that’s because most pitcher’s have a small spread in talent on balls in play: they’ve been selected to make sure that the 75% of the time that they pitch to contact, that they don’t get blown away. So, they’ve already been selected based on BABIP, so they don’t necessarily selected based on K/PA.
This is why GB rates for pitchers stabilize very fast: pitchers are not selected based on being GB or FB pitchers.
(Of course, there are other things that you aren’t selected for, that doesn’t stabilize: that’s because it has very little value.)
Goalies in hockey are very tight in their save percentages because that’s the ONLY thing they can do. So, you have no choice but to see a tight range (on a per shot basis). However, since they face some 1500 shots per season, what is hidden on a rate basis is more visible on a volume basis.
***
Inspired by this article.
Monday, August 29, 2011
Of all the sports that could start its own league, it’s basketball.
Hockey would have a bit of a tough time, for one big reason: The Stanley Cup. The Cup is actually not owned by the NHL, seeing that it predates the NHL itself. Instead, it has trustees appointed. There was some “transfer” or “delegation” that occurred that makes the NHL defacto “owner”, but until a court actually rules on that, the ownership thing is in limbo. Nonetheless, players play for the Stanley Cup, and anything else will be considered second fiddle.
Setting that aside, then hockey and baseball would face similar challenges: venues and tradition. The NHL did face competition from WHA in the 1970s, and the WHA was able to compete for talent based on salary, and they did attract star player, notably the older Bobby Hull and Gordie Howe (but still very talented). And the NHL has a draft age of 20 (!), so they also attracted “underaged” players like Wayne Gretzky (well, there was no one like him, but they got him and all the underaged players).
And NFL has done a great job of making football about the “team” and not about players. It doesn’t take much for a Packers fan to turn its back on Brett Favre. What counts is the uniform, just like college football.
But basketball? You don’t have 20 or 25 or 45 man rosters. You’ve got 12 guys. And, the superstar has a far larger impact in basketball than the other sports. So, you can easily create a 12-team league made up of superstars. You don’t have the same sense of history and tradition that you do with NHL and MLB. You don’t have the same live-and-die attitude as you do with football.
And forcing players to get less than they think they deserve is just what led to the NHL/WHA competition for talent. There’s a certain level below which players will revolt.
***
Related article: Forbes.
Lacob reportedly paid $ 450 million for the Warriors. That franchise price only makes sense if LeBron James, Dwight Howard, Kobe Bryant, etc… come play his Warriors. If these players are all in a new league, Lacob will stand to lose much of his investment in the Warriors. And the same story will be repeated for the other 29 owners. Faced with potential loss of the one thing fans are willing to pay to see (i.e. elite basketball talent), one suspects the stand the owners are currently taking will crumble.
And when that happens… well, I still think the players and cities should form their own league. Either way, though, fans will once again get to see basketball played at the highest level in the world.
Thursday, August 25, 2011
Ozzie Guillen believes it’s Peter Bourjos. Others might think it’s Brett Gardner. The first white guy to break 10 seconds is Marian Woronin of Poland in 1984. The two other white guys to break 10 seconds are Patrick Johnson (Australia) in 2003 and Christophe Lemaitre of France last year and this year. Koji Ito of Japan hit 10 seconds on the dot (though, the dot would be 2 or 3 decimal places).
The only good thing about pointing out that Bourjos is fast and white is that it shines a light on the accomplishments of others, and on the statistical oddity of it all. If that’s what it’s limited to, then I don’t have a problem with Ozzie’s declaration.
There was a time as recently as the 1980s where white boxers would be labelled “The Great White Hope”, which, today, seems outrageously racist and ridiculous to say, but back then, not many batted an eye.
Tuesday, August 23, 2011
My article at Fangraphs. This is good for those who want to see replacement level discussed in a somewhat different manner.
Thursday, August 18, 2011
In response to a query:
Whether it’s FIP or ERA or any rate stat, the way to “scale” them the same when you have differing opportunities (IP or plate appearances, etc) is the same:
value = (metric minus baseline) x opportunity
In your case, with FIP (or ERA), we’d like to establish the baseline as something like 1 or 1.25 runs above the league average. So, if the league average is an ERA of 4.00, then our baseline is 5.00 or 5.25. In other words, we create the baseline as the “minimum acceptable level” of performance. Think of a bullpen guy getting a spot start, or a last minute AAA callup, etc.
For opportunity, you use IP, but let’s use IP/9 so that it’s in the form of “games”. This is especially useful since FIP and ERA are also in the form of games (runs per 9 IP).
So, as an example, let’s say you have one guy with a FIP of 3.50 and 180 innings, and you have someone else with a FIP of 2.50 and only 108 innings. What are their values?
We’ll presume a baseline of 5.00.
value = (5.00 - 3.50) x 180/9 = 30
It may not be readily obvious, but the unit here is runs, so we have a pitcher who provided 30 runs of value above the minimum acceptable level (what we normally call the replacement level, or readily available talent level).
For the other guy:
value = (5.00 - 2.50) x 108/9 = 30
As you can see, these two pitchers are equivalent.
Such is life when dealing with uncertainty levels. When we say Jeff Francoeur is a replacement-level player, we don’t know for certain that he is. It’s just our estimate that he is. According to fWAR, he was close to 0 wins from 2008-2010. According to rWAR, he was below 0 wins. In the three seasons from 2005-2007, fWAR had him at 8 wins and rWAR at 6 wins.
When you make your estimate, you have to use all his data, not just his last three seasons. Of course, the further back in time you go, the less weight you give. Roughly speaking, the last three years should get double the weight to the three seasons preceding those. For most players, you can just look at the last three years, and get on with it, because you’ll get a similar answer. But with some guys, like Frenchy, who shows an 8 win gap between the current 3 years and preceding 3 years, you need to consider it.
So, entering 2011, rather than thinking he is a replacement-level player, you’d see he averaged around 0 wins in the last three seasons, and a bit over 3 wins per 162 games in the three seasons preceding those (using fWAR). Now, his average is a bit over 1 wins. Furthermore, seeing that he’s young, his true talent is on an upward slope. Finally, given that he was given so much playing time, regression toward the mean demands that he be compared to a more select group of players. All of a sudden, a guy who can be reasonably considered a replacement-level players under one perspective can then be considered at least a 1 win player, if not 1.5 wins, entering 2011.
Now, with Frenchy’s above average season this year, that (enhanced) estimate entering 2011 will be even higher entering 2012.
If we consider the entirety of his career, weighting each successive season at 25% higher than the preceding one, then Frenchy is at 5.3 WAR in 2236 weighted PA, or 1.66 WAR per 700 PA. We bump that up a bit because of regression toward the mean (10% toward 2.25 league average, or, if you want to be fair, something like 2.5 or 2.75 toward guys who play all the time like Frenchy). So, he gets an extra 0.1 WAR, to put him at 1.76 WAR. And then a bump of his past performance because of his age, which would be say another 0.1 WAR. We’ve got him now at 1.86 WAR. We give him 85% playing time, and now our new official estimate of Frenchy’s talent level entering 2012 is 1.6 wins. And for 2013, it’s going to be say 1.4 wins.
So, 2012-2013 is 3 wins of value, which at 4.5MM$ per win is 13.5MM$. And that’s what Frenchy’s next extension was signed at. Congratulations to Frenchy and the Royals for signing a fair deal.
Well, fair, as best as our uncertainty level allows us to say.
Tuesday, August 16, 2011
Or: how much is 26 games worth in wins?
If you follow wins above average, then 26 games is worth a bit less than 1 win.
If you follow wins above replacement, then 26 games is worth a bit more than 1 win.
(The gap between the above two is 0.36 wins.)
If you believe that missing a game means that you should get zero credit for your replacement player, then 26 games is worth 2 wins.
My guess is that most people don’t really believe the replacement-level model. I think most people think of it that if you don’t play, then it’s like you are leaving your team shorthanded. And I think that’s why the games played is going to be a huge penalty against Victorino.
To be more clear: the replacement-level model essentially means that you are being given credit for the performance of a replacement-level-kind of player.
Friday, August 12, 2011
These are all the pitchers born since 1952 (i.e., after Blyleven), with at least 60 starts, through age 27. They are ordered by ERA+ from worst to… uh, not-worst. The general rule is that if you’ve been given 60 starts, you are not going to get many more. The more non-worst you are, the more chances you are given. That’s why you see near the non-worst point, guys with over 100 starts. Kyle Davies stands out as someone who has been given an enormous number of starts at such a poor performance level. You’ll also note that a good portion of the players were given a fair number of games in the bullpen. Basically: it’s not working out here, let’s try you over there. Kyle Davies stands out as someone who was kept in the starting rotation. At some point, scouting has to give in to empirical data: as much as a scout may say that Kyle Davies is a decent pitcher, we have to accept that maybe he’s not that good.
Of course, right behind Davies is Mike Scott. You kids may not remember him, but he was one of the best pitchers of his era (at some point in his career).
Let’s look at the top 10 in this list, and see what they did in the 4 years after their age 27 season. How much hope can we possibly give the Kyle Davies of the world? (I had to exclude pitchers who are still to young, so I went down to the #15 on the list below to get my top 10.)
Here we go:
Van Poppel: 282 IP, 108 ERA+ (almost all in relief)
Bowen: out of MLB
Mike Scott: 796 IP, 106 ERA+ (almost all as starter), and then continued on for a few more excellent seasons)
Knapp: out of MLB
Snyder: out of MLB
Scudder: out of MLB
Walk: 530 IP, 115 ERA+ (half games as starter), and continued his career beyond
Rupe: 10 more innings then out of MLB
Wright: out of MLB
Codiroli (*): 122 IP, 75 ERA+
(*) I followed MLB intently when I was a kid, knowing every player on every team (easier done when you collect baseball cards, and are in fantasy leagues). I do not remember this guy at all.
So, that’s what you have here: 20% chance of good success, 10% chance of being useful as back of the bullpen guy, and 70% chance of being out of MLB.
Note: replacement level is ERA+ of 75-80 as a starter, and 95 as a reliever.
Glove-slap: Eric.
Source Baseball-Reference.com:
Tuesday, August 09, 2011
The Detroit Tigers are 61-53. The Tampa Rays are one game back, at 60-54.
When the NHL expanded from six to twelve teams, the NHL new that to give the expansion team a fighting chance, they’d have to put all six expansion teams in the same division. By letting the original six and expansion six have limited games against each other, it guaranteed that the expansion division would “look” competitive. But, in terms of head-to-head, the original six decimated the expansion six.
MLB, while nowhere near to this extent, has this kind of issue with the AL East. That is, since the Rays play against slightly tougher competition than the Tigers, it makes it harder for the Rays to pile up wins.
One way to show this is to look at “common opponents”, and in the same proportion. For example, since the Tigers never played the Reds, Marlins, Astros, Brewers, and Cardinals, we throw out all those 21 games from the Rays. So, gone are the Rays’ 12-9 record from those games. At the same time, the Tigers never played the DBacks, Rox, Dodgers, Mets, Pirates, and Giants. So, out goes their 7-11 record.
So, against common opponents, we are now at 54-42 for Tigers and 48-45 for Rays.
However, the Rays faced the Orioles 12 times, while the Tigers faced them only three times. So, what we do is pro-rate the record of the Rays against the Orioles down to 3 games. So, instead of being 6-6 against the Orioles, we make them 1.5-1.5. We repeat this with all the common opponents.
With 75 “matching” games (out of the 114 they played), the Tigers end up with 41 wins, and the Rays end up with 37 wins.
When I started this, I expected the Rays to be ahead of the Tigers, reasoning that the Rays had the tougher opponents. As it stands, they did have slightly tougher common opponents. But the two things that conspire against the Rays:
a. among the teams that only one of them played against, the Rays did much better, but all those games get thrown out
b. when the Tigers and Rays went head-to-head, the Tigers won all three games
Now, obviously, we don’t want or need to throw games out. We would simply do a strength of schedule adjustment with all the teams. Once you do that though, you are doing an indirect approach. You are comparing the Rays against NL Central with the other teams (not Tigers) also against the NL Central. And then you are comparing the Tigers against the NL West with the other teams (not Rays) also against the NL West. With those common baselines now in place (i.e., rest of league against NL West and NL Central), we presume that we can fairly compare the Rays and Tigers indirectly.
We could do that. But I’m not doing that here. In the process I’ve laid out here, we’re doing a direct comparison. And under the direct comparison, the Tigers are 4 games ahead of the Rays.
(Glove-slap to Max for inspiring this.)
Note: as an example of why it’s unfair to ignore the indirect methodology: the Rays could have been 21-0 against the NL and the Tigers could have been 0-18 against the NL, and under the direct methodology, all those games would get thrown out.
Wednesday, June 29, 2011
This gives you the number of pitchers and nonpitchers that were on each team for each year. I counted any nonpitcher who had as many plate appearances as there were team games (so, these days, that means at least 162 PA). For pitchers, I counted anyone with at least 0.23 innings per team game (so, these days, that means at least 37 innings).
We see that since the new run environment (1993-present), the % of roster spots allocated to pitchers is almost exactly 50%.
The jump started occurring in 1990, likely due to specialization of relievers. Prior to then, the number of relievers used topped off at 11.3 per team. From 1951-1984, the number of pitchers used hovered from 9.7 to 10.9 per team.
As for nonpitchers, it’s been holding quite steady since 1973 (DH?). Since then, we’ve had 12.5 nonpitchers per team. Prior to that, since 1922, it was 11.7 nonpitchers.
year Pitchers NonPitchers Players % Pitchers
1916 8.1 11.1 19.2 42%
1917 7.4 10.9 18.3 41%
1918 7.6 10.9 18.6 41%
1919 7.6 10.6 18.2 42%
1920 7.4 10.8 18.3 41%
1921 7.8 10.9 18.8 42%
1922 7.9 11.3 19.2 41%
1923 7.8 11.3 19.1 41%
1924 8.4 11.6 20.0 42%
1925 8.4 12.3 20.7 41%
1926 8.4 11.5 19.9 42%
1927 8.3 11.7 20.0 42%
1928 8.6 11.7 20.3 42%
1929 8.5 11.2 19.7 43%
1930 8.4 11.9 20.3 42%
1931 8.6 11.8 20.4 42%
1932 8.4 11.3 19.8 43%
1933 8.9 10.8 19.8 45%
1934 8.5 11.6 20.1 42%
1935 8.8 11.5 20.3 43%
1936 8.8 10.8 19.6 45%
1937 8.9 11.1 19.9 45%
1938 8.9 11.0 19.9 45%
1939 9.1 12.0 21.1 43%
1940 9.4 11.1 20.4 46%
1941 9.5 11.5 21.0 45%
1942 9.6 11.6 21.1 45%
1943 9.3 11.4 20.8 45%
1944 8.9 11.6 20.6 43%
1945 9.7 12.0 21.7 45%
1946 10.3 12.6 22.8 45%
1947 9.9 11.6 21.5 46%
1948 9.5 12.1 21.6 44%
1949 9.3 12.6 21.9 43%
1950 9.4 11.6 21.0 45%
1951 9.7 12.4 22.1 44%
1952 10.2 11.5 21.7 47%
1953 9.8 11.4 21.3 46%
1954 9.8 11.4 21.3 46%
1955 10.4 11.8 22.3 47%
1956 10.1 11.7 21.8 46%
1957 10.2 12.1 22.3 46%
1958 10.2 12.3 22.4 45%
1959 9.7 11.9 21.6 45%
1960 10.1 11.8 21.9 46%
1961 10.0 11.8 21.8 46%
1962 10.6 11.4 22.0 48%
1963 10.1 12.1 22.2 45%
1964 10.4 11.8 22.2 47%
1965 10.3 12.0 22.3 46%
1966 10.6 11.7 22.3 47%
1967 10.8 11.7 22.5 48%
1968 10.2 12.1 22.3 46%
1969 10.2 12.1 22.3 46%
1970 10.2 11.8 22.0 46%
1971 10.1 12.0 22.1 46%
1972 9.9 11.5 21.4 46%
1973 9.8 12.2 21.9 44%
1974 9.9 12.2 22.1 45%
1975 10.2 12.5 22.7 45%
1976 10.1 12.5 22.6 45%
1977 10.2 12.2 22.3 46%
1978 10.2 12.2 22.4 46%
1979 10.7 12.2 22.9 47%
1980 10.2 12.9 23.2 44%
1981 10.7 11.7 22.3 48%
1982 10.6 12.2 22.7 47%
1983 10.8 12.3 23.1 47%
1984 10.9 12.5 23.4 47%
1985 11.3 12.4 23.7 48%
1986 11.0 12.9 23.9 46%
1987 11.0 12.7 23.7 47%
1988 10.7 12.3 22.9 46%
1989 11.3 12.9 24.2 47%
1990 11.8 12.6 24.4 48%
1991 11.7 13.0 24.6 47%
1992 12.1 12.8 24.9 49%
1993 12.2 12.7 24.9 49%
1994 12.0 12.5 24.5 49%
1995 12.1 12.9 25.0 48%
1996 12.3 12.3 24.6 50%
1997 12.4 12.8 25.1 49%
1998 12.2 12.7 24.9 49%
1999 12.5 12.6 25.1 50%
2000 12.3 12.9 25.2 49%
2001 12.5 12.3 24.8 50%
2002 12.4 12.6 25.1 50%
2003 12.5 12.5 25.1 50%
2004 12.3 12.6 24.9 49%
2005 12.2 12.4 24.6 50%
2006 13.0 12.5 25.5 51%
2007 13.1 12.7 25.8 51%
2008 13.1 12.5 25.6 51%
2009 13.3 12.5 25.8 51%
2010 12.5 12.7 25.2 50%
Sunday, June 26, 2011
Phil finds it for WP and PB.
He also asks about it for fastball speeds, and I seem to remember seeing that data published somewhere. Anyone?
Saturday, June 25, 2011
Great stuff. The answer is 22% in the NHL.
I seem to remember doing something for MLB a year or two back, and I seem to remember getting an answer around 15% - 20% I think.
Basically, the best team in the league is about a 4:1 or 5:1 odds of winning that league’s championship. And that sounds about what Vegas puts the odds.
If someone has the Vegas odds for the last 10 years in all the sports, I’d like to see what those odds are.
This is for 2005-2010.
First column is performance against other NL teams. Second column is performance against all AL teams. Diff is the difference between the two. 13 of the 16 NL teams perform worse against AL teams than NL teams. Colorado however has a huge reverse-split. Presumably (speculatively?), AL teams are so disoriented playing at Coors that it confers a huge home field advantage to Colorado.
And the playoffs guarantees 4 teams in the NL and 4 teams in the AL every year? Ridiculous. I’d guarantee 3 teams from each league, and whichever league has the better interconference record gets to send 5 teams in all (with one of the 5 swapping to the other league). Ridiculous? Just trying to be ridiculous but fair, rather than ridiculous and unfair.
Inside Outside Diff
0.486 0.581 0.095 COL
0.422 0.444 0.022 WSN
0.497 0.505 0.008 FLA
0.483 0.462 -0.021 CIN
0.502 0.478 -0.024 MIL
0.523 0.495 -0.028 NYM
0.543 0.511 -0.032 STL
0.495 0.427 -0.068 SFG
0.406 0.333 -0.073 PIT
0.524 0.444 -0.080 ATL
0.481 0.398 -0.083 ARI
0.508 0.422 -0.086 CHC
0.505 0.394 -0.111 HOU
0.513 0.385 -0.128 SDP
0.577 0.404 -0.173 PHI
0.533 0.344 -0.189 LAD
Thursday, June 23, 2011
Good stuff from what Jeremy did here.
***
As an aside, the paragraph with this is used as background reference:
There is one skill among one group of players that has likely held constant over time—pitcher hitting. Because pitchers are selected independent of their hitting ability, any variation in performance in pitchers as hitters on a league-wide basis can be attributed to pitcher quality.
It’s not necessarily true. A few years ago, I showed how pitchers-as-batters hitting talent changed for those pitchers born before/after 1952/53. Not coincidentally, the DH rule would affect pitchers born close to those birth years. But, that’s an aside, because it doesn’t seem Jeremy actually did something with this.
Wednesday, June 15, 2011
In an ESPN article last year, I showed each team’s payroll relative to league average over a period of 12 seasons, along with how many times they won 89 games (a playoff-level number).
If you were to have two conferences split by payroll, you’d put the top 8 spending teams in one conference, and the other 22 teams in the other conference. That’s because you have a similar number of playoff-level teams as a result (48 in one and 52 in the other, over the 12 year period).
There’s a major, enormous inequity based on market size (represented as a proxy only for illustrative purposes by payroll). All the other stuff, the unbalanced schedule, the number of teams per division, the byes, the inter-conference games… all that stuff is tiny compared to the enormous structural advantage that simply existing in a large market confers to teams.
Payroll Index 89 Wins Team
214% 11 NYY
153% 9 BOS
143% 4 NYM
137% 3 LAD
130% 7 ATL
121% 3 CHC
118% 4 SEA
115% 7 ANA
114% 6 STL
111% 2 TEX
109% 0 BAL
106% 5 SFG
103% 5 HOU
103% 4 ARI
101% 3 PHI
99% 6 CLE
97% 3 CHW
95% 1 DET
94% 0 TOR
90% 2 COL
79% 1 SDP
76% 1 CIN
72% 1 MIL
67% 6 OAK
64% 4 MIN
64% 0 KCR
58% 1 TBD
57% 0 WSN
57% 0 PIT
52% 1 FLA
Monday, June 13, 2011
Derek does exactly (one of the way of) what I do. I don’t know that I actually get the same results, but, the process is bang-on.
Stabilizes Years Stat Denominator
100 0.2 K PA-IBB-HBP
168 0.3 UIBB PA-IBB-HBP
253 0.4 IBB PA
501 0.8 HBP PA-IBB
959 2.1 1B PA-HBP-K-BB-HR-ROE
833 1.8 2B+3B PA-HBP-K-BB-HR-ROE
48 1.5 2B 2B+3B
48 1.5 3B 2B+3B
1126 2.4 1B+2B+3B (BABIP) PA-HBP-K-BB-HR-ROE
143 0.3 HR PA-K-BB-HBP
62 0.5 HR (HR/FB) OF FB [MLBAM]
65 0.5 HR (HR/FB) OF FB [RS]
109 0.2 GB [MLBAM] GB+OF+IF+LD
116 0.2 GB [RS] GB+OF+IF+LD
182 0.4 OF FB [MLBAM] GB+OF+IF+LD
189 0.4 OF FB [RS] GB+OF+IF+LD
194 0.4 IF FB [MLBAM] GB+OF+IF+LD
233 0.5 IF FB [RS] GB+OF+IF+LD
795 1.7 LD [MLBAM] GB+OF+IF+LD
979 2.1 LD [RS] GB+OF+IF+LD
Inconclusive* SB% SB+CS
39 0.3 SBA% 1B+UIBB+HBP+ROE+FC
UPDATE: For pitchers:
Stabilizes Years Stat Denominator
126 0.2 K PA-IBB-HBP
303 0.5 UIBB PA-IBB-HBP
943 1.5 IBB PA
1346 2.1 HBP PA-IBB
3893 8.4 1B PA-HBP-K-BB-HR-ROE
2305 5 2B PA-HBP-K-BB-HR-ROE
4977 10.7 3B PA-HBP-K-BB-HR-ROE
1882 4 2B+3B PA-HBP-K-BB-HR-ROE
351 11 2B 2B+3B
351 11 3B 2B+3B
3729 8 1B+2B+3B (BABIP) PA-HBP-K-BB-HR-ROE
1271 2.7 HR PA-K-BB-HBP
1239 9.4 HR (HR/FB) OF FB [MLBAM]
105 0.2 GB [MLBAM] GB+OF+IF+LD
205 0.4 OF FB [MLBAM] GB+OF+IF+LD
288 0.6 IF FB [MLBAM] GB+OF+IF+LD
2026 4.3 LD [MLBAM] GB+OF+IF+LD
36 2.3 SB SB+CS
161 1.2 SBA 1B+UIBB+HBP+ROE+FC
Wednesday, June 01, 2011
Poz hates it even more than I do. He has a Pozcast with Bill James, where James talks about it (presumably it’s what we’ve already heard from him, but Poz doesn’t go into it, because he’s pushing the Pozcast… and I’d gladly oblige to listen to it, but I, like most people, am at the office, and I, like many, don’t have a speaker, and if I did, I, like several, would be told to not play that at the office; and I’d read his reader comments, but blogspot is also blocked at the office!).
I agree with Poz. It’s like completely removing Wayne Gretzky and Michael Jordan from the ice/court, by giving a lesser shooter an extra two feet around him to make the play. At least in football and soccer, when you double-team, the player is still on the field, and that player is still sometimes involved in the play. To emasculate Gretzky and Jordan is extremely anti-sport.
Anyway, I’ve talked about this in the past:
Vladimir Guerrero is up at bat. He is prepared to swing at anything close to the plate. Anything! And still, teams will intentionally walk him. Was there a more tension-reducing sight than when Barry Bonds was coming to the plate with 1B open? This is the complete opposite of what should have happened, and was not what the fans paid to see.
The rule is simple: Any 4-0 walk, intentional or not, results in a two-base penalty. If you have a runner on 2B, the 4-0 walk gets you runners on 1B and 3B. If you have a runner on 3B, then it’s guys on 2B and 3B. And, with runners on 2B and 3B, the batter goes to 1B, the runner on 2B stays put, and the runner on 3B scores.
Under this scenario, how often would a pitcher not give the batter at least one strike? Again, fans win, and the players go back to giving us action and tension.
***
Poz also asks how many walks would a batter need to get if all he did was walk and strikeout (and is a terrible fielder). This is why we have wOBA. It answers the question pretty clearly. Since you need to be a league average hitter to be a DH, that would mean that you need a wOBA of around 0.333 (or whatever the league average OBP is). And with the coefficient of a walk being 0.72 (and the strikeout would be around -.02), then you solve for this:
OBP * 0.72 + (1-OBP) * -.02 = .333
That gives you OBP = .477
So, it’s pretty close to .500. At .500 OBP (that is, one walk for every strikeout), your wOBA is .350. So, you can get about 1.10 strikeouts per walk, and still be breakeven.
Now, even easier than wOBA is Linear Weights, where the run value of a walk is around .32 and a strikeout is around .29 (it depends on the run environment). And .32/.29 = 1.10. This is why I love Linear Weights.
And, if you remember, I talked about K-BB per PA for a pitcher as a great way to measure a pitcher. And a pitcher who strikes out as many as he walks is a replacement level pitcher.
Therefore, to tie it all in, a batter needs to walk more than he strikes out and a pitcher needs to strike out more than he walks. In order to not follow that, they need to bring more to the table. A batter can bring power or a glove or his legs. A pitcher? Well, it’s extremely difficult for a pitcher to bring ANYTHING else to the table. He’s going to have to be an extreme GB hitter to limit “power” (i.e., few HR), or be an extreme pickoff pitcher to limit “legs”, or be able to control balls in play to an extreme extent to leverage his fielders “gloves”.
Monday, May 23, 2011
I’m not much an NBA fan. I know the stars, and that’s pretty much it. If we make a list of the best ballplayers in my generation (Bird/Magic), and going with nothing more than guessing, I’ll say, in random order: Bird, Magic, Isaiah, Stockton, Karl Malone, Jordan, Pippen, Drexler, Barkley, Iverson, Lebron, Shaq, Kobe, Duncan, Robinson, Garnett, Hakeem, Novitzki, Nash, Kidd. Again, no idea how good that list is. I’m just throwing out names I’ve heard. Let’s see, that’s… 20 names. Good, makes dividing easy. Of those, Bird, Stockton, Novitzki, and Nash are white. That’s 20%. I seem to remember that 25% of all NBA players are non-Black. So, I think it’s safe to say that based on my priors, that we should expect about 75% of the greatest basketball players to be Black.
But, before whatever the color line is in the NBA, we’re going to get a ton of non-Blacks in the discussion as greatest ever. This is of course terrible unfortunate and heavily biased against the great Black players who couldn’t make it in the NBA. And so we have Kareem highlighting the greatest basketball team that we don’t know (amidst his tame statue controversy).
***
Since baseball doesn’t have that level of white/black imbalance, the problem isn’t as bad, but it still exists. Which is why I pretty much draw the line at players born since 1931 (Mays/Mantle). That 1931 birth line is pretty useful for me, because you have to go back to players born in 1926, or maybe 1918, to find superstars. And starting with players born since 1931, you were getting tons of great ballplayers born (at double the pace of pre-1931 born players). Whether that’s because of opportunities given to non-Whites, or opportunities based on expansion, or opportunities based on MLB becoming a more viable vocation, who knows. But, it does save me from having a disproportionate number of white players being the greatest ever by drawing that line.
Saturday, May 21, 2011
The fewer the runs that score, the more chance that the game will still be tied after 9 innings. Seems intuitive enough. Here’s how the relationship looks:
from Amazon
Recent comments
Older comments
Page 2 of 344 pages < 1 2 3 4 > Last »Complete Archive – By Category
Complete Archive – By Date