Ten Things I Didn’t Know Last Week

It’s the weather

In my weekly Heater article, I reviewed the current status of run scoring. In a nutshell, runs are down because home runs are down. If teams keep scoring at the current pace (admittedly, pretty much a meaningless presumption), this will be the lowest scoring year in the past 15 seasons.

Chris Constancio has already talked about the impact of the weather on scoring, but I wanted a little more info. So I approached Mr. Home Run (no, not Barry Bonds. I’m talking about Greg Rybarczyk of Hit Tracker) and asked for some insight. As usual, his insight was pretty darn good:

In 2006, the average April homer picked up 2.1 feet from wind, and lost 1.6 feet from temperature, for a net weather impact of +0.5 feet. In 2007, the average April homer picked up 1.9 feet from the wind and lost 3.7 feet from temperature, for a net of -1.8 feet.

Counting altitude (which was a wash), the average homer in April 2007 flew 2.7 feet less, which means a few long flies that might have made it in 2006 did not in 2007. It’s hard to say for sure how many, but enough to take a few runs off the board for sure.

Note that Greg’s numbers are estimates with a decent degree of uncertainty. I looked at this trend in a little more detail in Heater, and concluded that it’s too early to say that weather alone has been the reason for the lack of home runs.

The new First Inning

Speaking of Chris Constancio and great websites, First Inning has been upgraded with a bunch of terrific new features. First Inning is a minor league/prospects website that tracks the stats of every minor league player, with a number of site specialties thrown in. For instance:

Just like voting in Chicago, you should visit Chris’ website early and often.

They just don’t run the bases like they used to

Rob Neyer recently complained about sluggardly sluggers (ESPN subscription required), guys who just jog down to first base on groundballs. Although he defends these sluggers by saying a possible hamstring injury isn’t worth losing a top batter, he has a point: As a whole, ballplayers don’t run like they used to.

John Walsh pointed this out in the 2007 Hardball Times Annual, saying…

Baserunners have gotten more conservative over the period of 1957-2005. This makes sense, given the general increase in offense over that period; better not to get thrown out when the next guy up has a good chance to hit a home run. Going back to our runner going from first to third on a single to right: In 1960 he tried for the extra base 59% of the time and was thrown out 4% (compared to 44% and 2% in 2000).

I did a little research of my own into the subject, thanks to John Jarvis’ fantastic website (where has has analyzed all of the Retrosheet files from 1957 to 2005) and found that runners are definitely not taking the extra base as often as they used to.


It’s true what they say: Baseball is getting more stationary, and duller, but it’s a trend that started in the 1960s, if not earlier. This, despite the fact that people are walking faster than they used to.

When teams score, in WPA terms

I’m a dangerous man with a little Win Probability data. And thanks to Fangraphs, we now have all the Win Probability data we could hope for (the latest addition: WPA logs for individual players). So I used it to answer a random question that popped into my head: Do teams score a lot when they’re ahead by a lot? I looked at run scoring by WPA grouping (from 0% Win Probability to 100%, in groups of 20% increments), and here’s what I found:

WPA          Runs    Pct.
0.0-0.2      554     13%
0.2-0.4      530     13%
0.4-0.6     1065     26%
0.6-0.8      972     24%
0.8-1.0     1002     24%
Total       4123    100%

As you can see, teams score more runs when they’re close or ahead, which makes sense. Teams that are losing are probably facing a pretty good pitcher, so they aren’t going to score a lot of runs to begin with. But, as I thought about it, I realized that there are probably a different number of plays for each WPA situation. And indeed there are (home team win probability only):

WPA         Plays    Pct.
0.0-0.2     6876     19%
0.2-0.4     5749     16%
0.4-0.6    11037     30%
0.6-0.8     5853     16%
0.8-1.0     6977     19%
Total      36492    100%

Look at that. The distribution of plays is symmetrical. What’s even more interesting is that 30% of all plays occur within 40% and 60%, only 16% at 60% to 80% (or 20% to 40%) with an uptick to 19% for the last group. So individual plays within a game are more likely to occur in really close or runaway situations, not so much in between.

Anyway, to finish the analysis, I divided runs scored per 100 plays to find the “true rate” at which teams score runs.

An Angell at Spring Training
For decades, Roger Angell's writing has warmed us to the romance of the new season.
WPA      Runs/Play
0.0-0.2     8.06
0.2-0.4     9.22
0.4-0.6     9.65
0.6-0.8    16.61
0.8-1.0    14.36
Total      11.30

This is a pretty good table, I think. Run scoring rises as a team’s winning edge increases. Plus, there appears to be some “let-up” when teams reach an 80% probability of winning, probably due to substitutions or, well, sluggardly sluggers.

Who’s responsible for Pythagorean variances

That’s a mouthful, but you may know about the Pythagorean record, which simply takes a team’s record of scoring and allowing runs and turns that into a projected won/loss record. It works pretty well, and I’ve always been fascinated by teams that substantially vary from their Pythagorean record, wondering “What’s going on?”

Last year, I developed a technique using WPA to assign responsibility for Pythagorean variances to batting and pitching. The notion is simple: teams can get more out of their run differential by winning more close games, or they can get less by scoring most of their runs in blowout games. Those are just two examples of any number of things that can cause a Pythagorean variance.

Anyway, I have applied the analysis to this year’s records (through Tuesday’s games) and found the following differences:

Team         Bat   Pitch   Tot
ATL          1.8     1.7     3
DET          0.6     2.5     3
MIL          0.3     2.2     2
TB           1.5     0.7     2
CHA          3.3    -1.3     2
ARI          1.1     0.9     2
CLE          2.3    -0.3     2
SEA          2.2    -0.4     2
STL          1.9    -0.3     2
PIT         -0.5     1.8     1
LAN         -0.4     1.7     1
COL          1.1    -0.2     1
LAA         -0.3     1.0     1
NYN          0.7     0.0     1
BOS          0.9    -0.4     0
HOU          0.2    -0.8    -1
WAS         -1.7     1.0    -1
BAL         -1.1     0.3    -1
FLA         -0.9    -0.2    -1
TOR          0.0    -1.2    -1
SD          -2.8     1.5    -1
SF          -0.1    -1.4    -1
TEX         -1.4    -0.1    -1
MIN         -1.7     0.3    -1
PHI          0.7    -2.3    -2
OAK         -1.5    -0.4    -2
KC          -2.2     0.0    -2
NYA          0.5    -3.7    -3
CIN         -2.4    -1.0    -3
CHN         -2.0    -1.5    -3

Among teams with positive variances, Detroit and Milwaukee can thank their bullpens while Tampa Bay can thank its batters. The White Sox’ batters have contributed quite a bit to their record (which is interesting, because they’ve hit very poorly in clutch situations), but their pitchers have let them down a bit. Conversely, the Padres’ batters haven’t been contributing much at all, but that Padre bullpen has been dynamite.

On the negative side, most teams at the end of the list can primarily blame their batters for their negative variances, except for the Yankees and the Phillies, who can point the finger at their pitching (particularly their bullpens, once again).

How the White Sox have done it

So, I was intrigued by the White Sox. They don’t bat well in high-leverage situations, they are last in the league in batting with runners in scoring position, and yet their bats have been positive contributors to Chicago’s Pythagorean variance.

Well, that’s another mouthful, isn’t it? What I’m trying to say is the White Sox’ bats have contributed more to the team’s wins than you’d expect, given the total number of runs they scored, and despite not batting well in the clutch. That’s still a mouthful, but it’s the best I can do. And I was intrigued. How can this be? By just about any definition of “clutch,” the White Sox aren’t it.

So I went back and looked at the number of runs each team has scored by WPA, and guess what? The White Sox score relatively more runs when WPA is between 0.4 and 0.6 (in other words, when the game is close) than any other major league team. Here are the four top teams in scoring when games are close, using an index where 1.00 is equal to the major league average (given the number of runs each team has actually scored):

Team     Index
CHA       1.31
ATL       1.28
PHI       1.23
BAL       1.22
ARI       1.19

How has this happened? Well, it turns out that the White Sox have scored a lot of runs in the fourth, fifth and sixth innings, when the score is relatively close. They haven’t necessarily batted well in the late innings, but they’ve done it midway through their games, when it matters more than in the beginning.

This simply means that commonly used “splits,” such as batting with runners in scoring position, may not tell you everything you need to know. WPA helps fill the gap.

The Devil Rays must fix their infield defense

The Devil Rays are finally becoming the interesting team to watch that we all thought they would be. Although they’re 16-22, they have a number of fine young players in the field, and a lot of good young arms on the way, too.

Jamie Shields may have passed Scott Kazmir as the staff ace and Rotoworld’s Rotation Report (which you can access by becoming a premium member), feels that the Devil Rays may have a future star rotation, with Jeff Niemann, Wade Davis and Jason Hammel soon joining Kazmir and Shields in the rotation. And don’t overlook Andy Sonnanstine; Rays Index hasn’t.

But if there is one thing young pitchers need for their confidence, it’s good fielding behind them. And the Devil Rays’ infield defense is horrid. Compared to the average team, their infield has allowed 29 groundballs to go through for base hits, a staggering amount. If they had an average infield, they’d probably have two more wins on the ledger, and they’d be giving their pitchers a much-needed confidence boost.

The record for most singles in a game

A couple of weeks ago, Minnesota hit 12 singles against the Devil Rays with no extra-base hits at all. That struck me as extraordinary, so I decided to look up the record for most singles, without any extra-base hits, in a game. Thanks to Baseball Reference’s Play Index, I was able to find the answer pretty easily.

Boy, was I off. The major league record for most singles in a game without any doubles, triples or home runs is 22, spanked by the Dodgers on June 3, 1988. The Dodgers beat the Reds, 13-5. The second-most singles ever hit without an extra-base hit was 19, also by the Dodgers, on May 24, 1973. Amazingly, they scored only three runs and lost to the Mets, 7-3 in 19 innings.

There have been 271 games with more than 12 singles, and no extra-base hits, since 1957.

Ichiro is on a record pace

Through 34 games, Ichiro Suzuki has tallied 116 putouts in center field for the Mariners. If he maintains that pace over 162 games, he’ll total 552 putouts. The current major league record is 547, set by the Cardinals’ Taylor Douthit in 1928.

Can Ichiro maintain the pace and break the record? Perhaps. Detect-O-Vision takes a closer look.

Milwaukee fans are excited

As well they should be. Become a real Brewers fan and vow to pee in your pants if the Brewers win it all (hat tip: Ballhype and Deadspin).

And even if you’ve heard of that, have you seen Michael Tejera’s wild pitch?

References & Resources
Note: I corrected the info about most singles in a game after this article was originally published. I had originally said that the Mets had the second-most singles in a game without an extra-base hit. It was the Dodgers.

Dave Studeman was called a "national treasure" by Rob Neyer. Seriously. Follow his sporadic tweets @dastudes.

Comments are closed.