Players reaching base in the most games in 2010
Posted by Andy on March 31, 2011
We had a little discussion on a previous post about Derek Jeter reaching base in 87% of his games in 2010.
I went ahead and calculated this same percentage for every single player, excluding pitchers.
Here are the leaders in highest percentage of games reaching base in 2010, minimum 50 games played:
Logan Morrison 0.919 21 Prince Fielder 0.894 11 Evan Longoria 0.894 47 Miguel Cabrera 0.893 3 Robinson Cano 0.881 32 Joey Votto 0.880 2 Justin Morneau 0.877 1 Shin-Soo Choo 0.875 12 Ian Kinsler 0.874 31 Nick Markakis 0.869 49 Billy Butler 0.867 25 Derek Jeter 0.866 143 Josh Hamilton 0.865 7 Joe Mauer 0.861 10 Chase Utley 0.861 27 Matt Holliday 0.861 22 A McCutchen 0.857 60 Bobby Abreu 0.857 94 Albert Pujols 0.855 4 Mark Teixeira 0.854 58
Logan Morrison's agent should be reading this for when he becomes arbitration eligible.
The 3rd column is the rank in OBP for the season for each guy (again among players appearing in a minimum of 50 games as non-pitchers.)
Amazing that Derek Jeter is so low!
Now here are the 20 worst players in terms of fraction of games reaching base, minimum 50 games played:
Delwyn Young 0.409 187 N Schierholtz 0.409 108 Nick Stavinoha 0.405 52 Joaquin Arias 0.403 30 Brooks Conrad 0.398 198 Matt Stairs 0.397 110 Jason Bourgeois 0.391 70 Ramiro Pena 0.388 11 Greg Dobbs 0.386 14 Dewayne Wise 0.385 46 Augie Ojeda 0.373 17 Brandon Wood 0.370 1 Jesus Feliciano 0.370 35 Garret Anderson 0.363 2 Ross Gload 0.362 218 Travis Ishikawa 0.353 175 Alb Gonzalez 0.342 37 B Lillibridge 0.328 13 Cory Sullivan 0.246 16 Chris Carter 0.200 161
The third column here is the ranking among the worst OBP values for 2010, again for the same criteria.
These two tables show that there's quite a bit of variation in terms of how guys spread out their times on base. It's also true some of the guys on the second list are hurt by being pinch-hitters. Ross Gload, for example, had 79 games with just one appearance, and even though he had a .328 OBP, that means he had a ton of games with no times on base despite not being an awful hitter. That's how he gets on this "worst" list despite having only the 218th worst OBP.
March 31st, 2011 at 9:30 am
Casey McGehee thanks Prince for getting on base so often, I'm sure.
March 31st, 2011 at 9:59 am
Forgive my stupidity, but what do the numbers in the second column represent?
March 31st, 2011 at 10:07 am
So Jeter is great at getting on base exactly once a game??
March 31st, 2011 at 10:16 am
Jeter got on in 86.6% of games...where are you getting 1?
March 31st, 2011 at 10:23 am
@2 - the numbers in the second column represent the percentage of games in which Jeter reached base at least one time. He played 157 games and reached base via a hit, walk, or HBP in 136 of those games.
Andy - how many of those guys in the second list had 100 games or 400 PAs or something more than 50 games? It would be interesting to see how often the worst "regulars" reach base. For instance, in 1992 Jose Lind played in 135 games and had 506 PAs. He reached base in 92 of 135 games, around 68% (he had a wonderful .275 OBP that season). You mention that Gload appeared a lot as a pinch hitter, but I think the same is true for a lot of those guys on the bottom. Schierholtz, for instance, had 21 games in which he didn't even receive a PA (I'm guessing defensive replacement). Is it easy enough to generate a list with a 100 game minimum and maybe 3.5 PAs per game?
I actually think the percentage of games in which the worst offensive players who are regulars reach base is higher than most people would think. I'm guessing most people would see Lind's .275 OBP and wouldn't think he reached base safely via hit, walk, or HBP in 68% of his games.
March 31st, 2011 at 10:27 am
Artie, I could look at some of that, but I don't have the data file with me right now.
As for Lind, it's pretty easy to guestimate something like that. If he's got a .285 OBP and gets 4 PA in a game, then the odds that he DOESN'T reach base at all are (1-0.285) ^ 4, in other words he has a 71.5% chance of not getting on base in any given plate appearance, so the odds of not getting on in all 4 is .715 ^ 4, or 26.1%. The odds, therefore, that he gets on base at all are 73.9%, meaning we'd expect him to get on base in about 74% of his games, if he truly got 4 PAs each game (and if they were all independent events, which they were not.) Lind actually averaged 3.7 PAs per game, so it was easier for him to register the occasional no-times-on-base game.
March 31st, 2011 at 11:35 am
What would a historic list look like? I believe DiMaggio followed up his 56 game hitting streak with another long one and got on base in the game in between. That'd give him close to 100 games right there, plus I'd guess he did pretty well in the other games. Has anyone ever reached in 100% of games played? Even with a low threshhold?
March 31st, 2011 at 11:37 am
I guess I shouldn't say I BELIEVE that happened. I've heard it said that he had two long hitting streaks with just one game in between in which he got a walk. It came up a few years ago when Jeter had a long consecutive games-on-base streak and people conjectured if it was the longest ever (probably before BR existed or at least hit the quasi-mainstream) and people immediately knew it wasn't longest because of the DiMaggio streak.
March 31st, 2011 at 11:40 am
It's easy to figure, BSK. Go to DiMaggio's 1941 game logs.
He got on base 74 straight games.
March 31st, 2011 at 12:48 pm
On Jeter's % of games reached rank being much higher than his OBP. He
bats at the top of the order for a high scoring team. I bet he ranks very high in PA/gm. Of course having more PA's is very helpful in this stat. (As Joltin' Joe's low BB% made it more likely he would have the long base hit streak than Splendid Splinter with his higher BA but many PA's were BB's and not AB's.)
March 31st, 2011 at 12:58 pm
As of 2006 Ted Williams had the longest consecutive on- base streak, 84 in 1949. Joe DiMaggio is second with 74 in 1941.
March 31st, 2011 at 1:38 pm
Jeter ranked 2nd in MLB last year with 739 PA (157 Games) behind Rickie Weeks with 754 (160 Games). I didn't do any calculations so I don't know how he ranked in PA/G, but there isn't anyone in the vicinity with dramatically fewer games so he's getting a lot of opportunities.
March 31st, 2011 at 3:56 pm
Alright, Andy, if you want to take the EASY way out! :-p
Can you run this search for all-time leaders? I guess we'd be limited to the games we have play-by-play data or box scores for (I believe there are still some time periods we lack, no?), but it'd be curious to see if anyone really excelled in this.
March 31st, 2011 at 4:10 pm
Using Richard Chester's data, I looked through Ted Williams game log for that year. I counted 6 games where he recorded neither a hit nor a BB (I double-checked for HBP, which he didn't have either, because I wasn't sure if they counted towards this). He played in 155 games that year, meaning he reached base in 96% of games.
In 2004, Bonds had 6 no-hit, no-BB games out of 147, also for a 96% success rate (Williams edges him if you don't round). However, 2 of Bond's games were games in which he pinch hit and only recorded on PA. If you factor those out, you get 4 games out of 145 for a 97% success rate. If you take out all pinch hit/1PA games, he still rounds to 97%.
Ted Williams did not have any pinch-hit games that year.
There may be guys higher than that, since I only searched those two seasons (Williams based on Chester's research and Bonds based on his ungodly OBP that year), but I'd doubt it. Those are some absurd numbers when you think about it.
March 31st, 2011 at 4:20 pm
Babe Ruth, 1923, had 8 such games out of 152 for 95%. There was one game in which he started, got 1 PA, and left. I'm not sure why. If we exclude this, it doesn't really change anything.
March 31st, 2011 at 4:22 pm
Man, I really need to subscribe to the PI. Manually searching is just hard! Especially because there isn't really a way to ensure it. I mainly worked off guys with absurd OBPs and Times on Base. But a guy could reach base exactly 160 times with an OBP of 250 and, if he spread everything out, he would trump these guys. Obviously, it's not likely. But, I can't rule anyone out unless their Times on Base was a certain amount lower than their GPs. Fun stuff!
March 31st, 2011 at 6:41 pm
Andy - I agree that it's easy to calculate the percentage of games a guy should reach base given his OBP. I just don't think many people (perhaps this does not apply to people who read the blog, but others) would realize a guy with a .275 OBP gets on base in about 65%-75% of his games played. Even though if enough time is taken to think about it people should, because Lind was actually on base 137 times in 135 games in 1992. The perception is that a low on-base guy is NEVER on base and so people may believe that such a player goes many games in the season without ever reaching base. I would almost wonder if people would equate OBP with percentage of games a guy is on base in a season (which would be really, really understated for regular players). As in "Lind is only on base 27.5% of the time, so he's only on base in about 30% of his games" or even "Bonds is on base 60% of the time, so he's on base in 60% of his games".
BSK - I don't think you will find anyone higher than Bonds/Williams unless it's a Ruth season or a 19th century player (McGraw, Hamilton - someone with a high OBP). When Andy discussed this on the other post I checked Bonds' 2004 season and saw (as you mention) he only had 6 games where he did not reach base. That is incredible.
Of course, this does lead to an interesting question. Within in a season, given two guys with the same set of stats (for simplicity say the guys get on base 160 times in 160 games in the same number of PAs and they have the same amount of power), do you want the guy who is more consistent (gets on base exactly once a game) or less consistent (gets on base two times in 80 games and 0 times in the other 80)? Who helps you win more games?
March 31st, 2011 at 6:43 pm
I should have written "estimate the percentage of games" in the first line of my prior post instead of "calculate the percentage of games" since it really is an estimate.
March 31st, 2011 at 7:22 pm
http://www.baseball-reference.com/blog/archives/425
has Frank Thomas with a 90 game streak of reaching base
March 31st, 2011 at 8:43 pm
@5 Sorry...I meant to ask what do the numbers in the THIRD column represent, to the right of the percentages. For instance, what does the 143 next to Jeter's name represent?
March 31st, 2011 at 8:47 pm
Tim, I explained what the 3rd columns mean right in the original post. It's the ranking in OBP for all players in 2010, with certain caveats you can read.
McCombe35, first of all, FIST BUMP for pulling out a 3.5 year old post of mine...I can't believe I've been blogging here that long. I believe that post is position-specific, and Frank Thomas probably had some games in the middle of that streak where he played DH and his actually consecutive games streak wasn't nearly as long.
March 31st, 2011 at 9:25 pm
@21
That list is position specific. The first game of Thomas' streak was on 6-29-95. On July 30 did not reach base as a PH and DH.
March 31st, 2011 at 9:43 pm
@19
Unfortunately, there's a twist to that old post. That's 90 straight games *as a first baseman*. I checked and Frank Thomas had some on-base-less games at DH during that streak.
Plus, the play index only went back to 1957 back then. It would be interesting to see what that list looks like now that it contains all the big live-ball seasons (Williams, Dimaggio, Ruth, Hornsby, etc).
March 31st, 2011 at 9:44 pm
Man, seems like forever ago that the PI only went back to 1957. My how times have changed.
I'm really grateful to have been blogging here for so long--and thanks to al you readers who have stuck around for so long too.
March 31st, 2011 at 10:14 pm
Andy -- Nice move including the 3rd column (ranking in OBP). If I may add a bit of context, there were 404 qualifying players (50+ non-pitching games).
BSK -- I would never say that everyone should subscribe to the Play Index, but once you've said out loud, "I really need to subscribe" ... you really need to subscribe! (Then, watch the hours melt away....)
March 31st, 2011 at 10:25 pm
Andy-- That explains it. Thanks.
I think there was a Thomas season where he reached base in all but about 8 games. probably the short '94 season.
March 31st, 2011 at 10:30 pm
Pop quiz (no P-I searching -- that's only for me!):
Name the modern player who had the highest single-season average of PAs per game (4.92), as well as 3 of the top 12 seasons. Minimum 50 games to qualify, but I'll go ahead and tell you he played over 140 games each of those 3 years. And he wasn't from the PED era.
I freely admit that I did not know this before I went a-searchin', and while he is a very well-known player, he would not have been among my first 5 guesses.
March 31st, 2011 at 10:34 pm
I did notice that Game Log pages have longest hit streaks and on base streaks. It also talks about the guys most frequently driven it. It was fascinating seeing some of the numbers for Bonds in his otherworldly years (I think in '01 he drove himself in 73 times and the next highest guy was like 24 or something).
March 31st, 2011 at 10:37 pm
P.S. to #27 -- The same player holds the modern career record (min. 1,000 games) with 4.63 PAs per game.
March 31st, 2011 at 10:39 pm
BSK- guessing there were a few players that had more than 24 HRs that year.
March 31st, 2011 at 10:45 pm
Is it Earl Combs? Yankees scored a lot of runs and he batted leadoff.
March 31st, 2011 at 10:46 pm
maybe Joe Morgan?
March 31st, 2011 at 10:48 pm
@26
If I did my search correctly Thomas did not reach base in 9 games in 1995.
March 31st, 2011 at 10:52 pm
@33
sorry, ignore my post #33.
March 31st, 2011 at 10:53 pm
@33
out of 145 league leading games. (another stike shortened season which most seamed to forget)
That games reaching base % would be tough to beat.
March 31st, 2011 at 10:53 pm
Rats!
March 31st, 2011 at 11:12 pm
DavidRF -- Not Combs, but very good guess. The competition is tough; our man needs to be not only a leadoff hitter for a historically high-scoring offense, but (for the career record) he needs to hit the ground running and retire the minute he loses a step. Our man spent very little time on the bench before establishing himself as an everyday player as a rookie; at the other end, he went directly from 128 games played (and his 7th All-Star berth) to 3 games and retirement.
March 31st, 2011 at 11:13 pm
I figured out who John Autin is talking about. I cheated though. Its not Combs or Morgan (or Rose). Crosetti would have been a better guess, but its actually not a Yankee.
I will give away that it looks like he did a post-1900 search. Billy Hamilton is the all-time leader (4.7668 R/G career, 5.2636 in 1894). If you include the NA, there are a number of players from 1872-1873 because of the high context and shorter (yet still > 50G) seasons. Ross Barnes-1873 is way up at 5.6667.
I'll stay mum on the post-1900 answer so as not to spoil the fun.
March 31st, 2011 at 11:27 pm
do you want the guy who is more consistent (gets on base exactly once a game) or less consistent (gets on base two times in 80 games and 0 times in the other 80)? Who helps you win more games?
This is something I often wonder about and I'm sure I've asked here as well. I'm not sure how best to approach the question and I'm not aware of any studies on it. Whenever I hear about a batting streak, or that Smith has X multi-hit games this season, I think about it. I'd guess that the answer varies depending on the context (scoring levels, teammates), but I'm really not sure which is better under "normal" circumstances.
March 31st, 2011 at 11:30 pm
McComb-
You misunderstood. In 2001, Bonds had 137 RBIs. Of those, 73 were from driving himself in. 25 were from driving Rich Aurillia in. 12 were from driving Marvin Benard in. Everyone else was single digits. It's kind of crazy to think what his RBI numbers would have been like had people gotten on base in front of him and he had an opportunity to bat with runners on. Yea, yea, I know RBIs are a bad stat. But in terms of describing what DID happen, they are useful. And what happened that year was crazy on multiple accounts.
March 31st, 2011 at 11:35 pm
DavidRF -- Thanks for staying mum, but I'll give the answer since I'm toddling off to bed soon: The Little Professor, Dom DiMaggio.
BTW, I did say "modern player," a generally accepted shorthand for post-1900. Also, while Crosetti did lead off for some historically potent Yankee teams (and owns 2 of the top 12 seasons in PA/G), he is not close to the career record, due to a long decline phase as a part-timer and defensive replacement.
March 31st, 2011 at 11:54 pm
Johnny Twisto -- I have no means of answering that question either, but I'd think that another variable (besides team and league context) would be the level of OBP for the players being compared. For instance, in the example you quoted, both players are terrible (OBP of .250 assuming 4 PA/G), but I'd guess that the guy who bunches his times on base is more valuable, because I think the incremental value of a 2nd time on base in a game is greater than that of the 1st.
But if we compared two guys with .400 OBP, one who gets on 2 times in 5 PAs every game, while the other gets on 4 times in half the games and no times in the other half, then my gut tells me the steady guy has more value.
Just speculation, of course.
April 1st, 2011 at 12:33 am
@ bsk
oh OK, got it.. Thanks.
April 1st, 2011 at 12:34 am
I couldn't resist a quick-and-dirty study of the comparative WPA value of a steady 1-for-4 vs. alternating 2-for-4 and 0-for-4 (as discussed @39).
I took the last 300 batter games of the 2010 season (by date) in which the batter had exactly 1 time on base in exactly 4 PAs, copied it into
Excel and figured the average WPA. I did the same for the last 300 games of exactly 2 TOB in 4 PA, and for 0 TOB in 4 PA. The results:
1 time on base averaged -0.015 WPA per game.
0 and 2 combined averaged -0.042 WPA per game.
(0 times on base averaged -0.100 WPA;
2 times on base averaged 0.058 WPA.)
The incremental values of the 1st and 2nd times on were similar, but the 1st was greater. The net WPA value of the 1st time on base was 0.085, that of the 2nd was 0.073.
The blunt interpretation is that the steady 1-for-4 is more valuable, because of a higher average WPA. But I don't really know how to interpret the WPA numbers, especially when both averages are negative. Obviously, you can't win a game without some positive WPA.
There are a dozen other flaws in this study, obviously. 300 games for each category is a tiny sample, so is subject to abnormal variation in WPA effect for those given offensive events. I didn't control for total bases; one group might have hit a bunch of HRs. I didn't control for quality of team; it's easier to build up positive WPA on a good team. Etc.
April 1st, 2011 at 12:44 am
I do think one (or more) Frank Thomas seasons would rank pretty high on a list of % of games reaching base. for a full season. And really high with the PEDs left off.
April 1st, 2011 at 12:45 am
@42
Your gut, huh. I was actually watching Stephen Colbert when I read that. 🙂
There was a similar study (the old Drysdale/Pappas comparison) which was based on good years/bad years that found the opposite. I'm not sure how analogous that study is, though, because they were treated "pennants" as the success marking and it didn't really matter how bad the off years were.. Of course, you'd rather have your team win 100 games half the years and 60 games the other half as opposed to winning 80 games every year.
When its games instead of years, you can't toss out the off-games because they're presumably happening in the same season. I'm thinking that with linear weights that they'd come pretty close to cancelling out. I doubt its an exact tie though.
April 1st, 2011 at 6:36 am
JA-
When I was a kid, I always wondered whether it was better to go 1-4 with a guaranteed HR every game or 4-4 with 4 singles. I usually concluded it was the former of the two, since you were guaranteed a minimum of 162 RBIs and possibly more. As I learned more about advanced stats, I switched positions, since the former has a .250/.250/1.000/1.250 slashes and the latter has 1.000/1.000/1.000/2.000. I'm curious if there are even more advanced stats that would more conclusively indicate which is more ideal.
April 1st, 2011 at 10:00 am
Awesome discussion here, folks. It's given me some post ideas...coming up.
April 1st, 2011 at 10:06 am
BTW, by my count, BSK has now posted 3 different times that he "really needs to subscribe to the PI".
April 1st, 2011 at 10:31 am
I found the answer to BSK's question in #47 and the post will be up Monday morning. So as not to be such a huge tease, the answer is that 4-for-4 with 4 singles wins out. You'll probably be surprised to see how close or not close it is, though.
April 1st, 2011 at 10:39 am
Alright, alright, I'll take the plunge. For some reason, I thought it was more than just $36 for the year. Before I do, I'm just curious if this subscription works at all SR sites? If not, do you offer any bundle discounts? In for a dime, in for a dollar. And it's not like I need a job or a fiance or friends or anything anyway...
April 1st, 2011 at 10:42 am
Thanks, Andy! I'm excited to see how it works out. You'll satisify both 12-year-old me and 27-year-old me!
April 1st, 2011 at 10:44 am
I'll warn you right now that the post includes a dig at you for not being a PI subscriber...heh.
Just to be clear to everyone, I don't see any money from PI subscriptions, so don't think my pressuring is to try to line my own pockets.
I just double checked, and the PI-like features on all the other sites are free. They are not so exhaustive and detailed as the PI here on Baseball-Reference.com, but they are all free. (So, buying the PI subscription here doesn't get you anything that you can't already get for free otherwise, but it does encourage Sean to keep developing this site!)
April 1st, 2011 at 12:01 pm
JA @44,
The 2 games combined for the 2-4 and 0-4 add to -.042. So the per game average is -.021. I think you may have some selection bias(es) going on here. I am not sure that the pools of players in the 3 groups; 0-4, 1-4, 2-4 would be of the same quality. They may also not be facing opponents of the same quality. (Especially important when using WPA)
April 1st, 2011 at 12:24 pm
Andy-
I figured as much. I had actually intended to subscribe a while back and then got caught up in things. I've been meaning to cancel my ESPN Insider subscription anyway (if only to make them STOP sending me 'The Mag') and I can just put that money towards the PI.
Again, if I get fired from spending too much time on it, I'm insisting that BR hire me.
April 1st, 2011 at 12:24 pm
"I figured as much" is in response to realizing you don't have a financial incentive in my investment...
April 1st, 2011 at 12:35 pm
Kds -- Thanks for catching my error. The moral of the story: Quick-and-dirty studies conducted after one's bedtime tend to come out even dirtier than intended.
(Or, if you prefer The Sports Guy's line: "The moral of the story, as always: I'm an idiot.")
April 1st, 2011 at 12:38 pm
To those who've been following the issue that I "studied" @44, a question:
Is looking at WPA (whether average, mean, or whatever) a good way of looking at this question? Or would it be better to look at raw Runs & RBI?
April 1st, 2011 at 1:00 pm
JA-
I don't think that WPA is a good measure. WPA tells us what did happen. A guy could post an extremely high WPA or an extremely low WPA outing with a 2-4 performance. It doesn't tell us the value of a general WPA... only of that specific player's circumstance. I suppose with a large enough sample, we may be able to generalize, but I still don't know.
Are we sure there is even a difference between the two? It's not like we're comparing HRs hit in bases-loaded situations in close games to HRs hit in bases-empty situations in blowouts. It's like we're comparing HRs hit on Tuesdays to HRs hit on Thursdays.
I highly doubt there is any quantifiable difference between going 1-4 every day and 2-4 every other day. Not in any way that we can generalize.
April 1st, 2011 at 1:03 pm
"I always wondered whether it was better to go 1-4 with a guaranteed HR every game or 4-4 with 4 singles."
I'll be interested to see Andy's approach. The question can't be addressed without making some assumptions about the surrounding context -- the caliber of the rest of the lineup, the overall offensive level of the competition, and your team's run-prevention ability, to name a few.
To cite an extreme example: If the other 8 men in your lineup all hit like Dean Chance, then I believe you'd much rather have the HR (assuming our singles hitter doesn't steal 2nd & 3rd base every time). As the quality of the lineup goes up, the value of the singles hitter goes up relative to that of the HR hitter.
April 1st, 2011 at 1:10 pm
JA (btw I am very excited that my nickname for you has caught on!)
I didn't consider any of that stuff but rather took a WPA approach similar to your method, or rather to kds' corrected method. I'd think averaged over lots of games, it's a fairly good method.
April 1st, 2011 at 1:29 pm
@59, BSK -- You may well be right about that specific comparison. But within the theme, there is clearly some level of production at which it makes a big difference.
Consider two batters getting 4 PAs every game. Angelo goes 1 for 4 with a HR every game, while Bombo hits 4 HRs in one game and goes oh-fer the other three games.
I'm pretty sure Angelo is much more valuable, because the incremental value of Bombo's 3rd and 4th HRs is very small. No matter how many runs you score in one game, you can only get one win.
April 1st, 2011 at 1:41 pm
JA-
Great point. I guess there is no general "right" answer, as much as I'd like there to be. Again, these ruminations started when I was about 12. Bombo might be the more valuable one if teams averaged 20 runs a game while in most other circumstances, Angelo would be.
What would be ideal is to run the query taking into account every reasonable base/out/score scenario. So Bombo might be more valuable in a season where teams average an absurdly high amount of runs, but that environment might only come up in .1% of season and, as such, his value would be impacted accordingly. But if there was a way to simulate every situation and say, "In X% of scenarios, approach Y was better and by Z amount and in A% of scenarios, approach B was better and by C amount." Am I making sense?
April 1st, 2011 at 6:07 pm
With reference to Williams 1949:
5 of the 6 games he failed to reach base happened in June (Jun.2,3,7,26,30)
The other happened the 4th-to-last game of the season
Bonds 2004:
Failed to get a walk in only 26 of 147 games (24/145 subtracting the 1PA appearances)