10 July 2007

First Half Hitting Game Grades

by Joshua Worley

Which Dodger hitters are most likely to have a good game at the plate? Which are most likely to have a bad game? This isn't the same as asking which players have the best stats and worst stats, though we'd expect players with good stats to end up having more good games. We know Russell Martin is better than Juan Pierre, and yet Pierre has played some excellent games and Martin has played some bad games.

I took each Dodger with a significant number of multi-at-bat games and put their game logs into a spreadsheet where each game could be graded on an A-F scale. The grading scale was simple, based on times on base, with a bonus given for extra bases earned, such as from doubles, stolen bases, and home runs. Caught-stealings are taken out of times on base.

Here is the basic grading scale based on percentage of times reaching base by hit, walk, or hit by pitch:

F -- 0%
D -- 1%-25%
C -- 26%-50%
B -- 51%-75%
A -- 76%-100%

One grade boost for 1-2 extra bases
Two grade boost for 3+ extra bases

The only way to get an F is to never get on base, or perhaps to get on base once and then get caught stealing. Juan Pierre had a game where he never got on base by his own effort but had a stolen base to boost that game grade to a D. A game where a player gets only a walk or a single with no steals will be a D game. If a player reaches first base twice or has a double then it will likely be a C game. B games are something like 2-4 with a walk, or 2-5 with a double, or 1-4 with a home run.

If a player hits a home run in a game, he's guaranteed at least a B, because of the two grade boost for the 3 extra bases the home run provides, and the D grade guaranteed by getting on base once. If a player hits a home run and gets on base at least one other time he'll get an A. A player without power or speed will have a hard time picking up many A's, as he'll have to be nearly perfect in getting on base to get one.

It's a tough grading scale, which is appropriate, given that baseball is a tough game in which even the best players may have an O-fer. Here are the 12 Dodgers who have played the most and their grade distributions. Only games with at least two plate appearances are graded.

nameABCDFtotal
Pierre11728291388
Martin191718121682
Gonzalez161320221182
Kent131914191479
Furcal101421191478
Garciaparra71315251777
Ethier111218131973
Betemit56861237
Abreu25105628
Kemp5651623
Loney5248019
LaRoche2343214

As expected, every hitter ( with a few sample-size exceptions ) will play a significant number of games at every grade level. A slightly better than average hitter, such as Jeff Kent, will have a nearly symmetric grade distribution. It appears to take a very good hitter indeed to earn more of the good grades than the bad grades, at least the way I've set up the grading scales.


How do the Dodgers hitters rank in percentage of games in which an A was earned?

A-list

Loney 26%
Martin 23%
Kemp 22%
Gonzalez 20%
Kent 16%
Ethier 15%
LaRoche 14%
Betemit 14%
Furcal 13%
Pierre 13%
Garciaparra 9%
Abreu 7%

Martin delivers a special game almost a quarter of the time. Loney and Kemp, though in far fewer games, are right there with him. At the other end, excellent games from Nomar and Abreu are rare and very unexpected. Even Pierre does significantly better than these two. Nomar seems to have no power or speed left, which accounts for his poor showing here.




How do the Dodger hitters rank in percentage of combined A or B level games?

AB-list

Kemp 48%
Martin 44%
Kent 41%
Loney 37%
LaRoche 36%
Gonzalez 35%
Ethier 32%
Furcal 31%
Betemit 30%
Garciaparra 26%
Abreu 25%
Pierre 20%

Even the best hitters are going to have C, D or F games more than half the time. This just shows how truly awesome Kemp has been in limited playing time, with nearly half his games being excellent or very good. Loney and LaRoche also stand out here for me. Juan Pierre is down to last place.



How do the Dodger hitters rank in avoiding D or F level games? Part of being a good hitter is having those terrific games, but just as important is avoiding the awful games the drag down a team's offensive effort. A low percentage of D or F games and high ranking on this list are good.

DF-list

Kemp 30%
Martin 34%
LaRoche 36%
Abreu 39%
Gonzalez 40%
Kent 42%
Loney 42%
Furcal 42%
Ethier 44%
Pierre 48%
Betemit 49%
Garciaparra 55%

Kemp is a juggernaut! After a third on the first list, he's had two first place finishes among Dodger hitters. Martin only has a bad game a third of the time. He really is the best. ( In a subjective, non-specific way. ) On the other hand, Nomar has a bad game more than half the time.


How do the Dodger hitters rank in avoiding F level games? This is essentially the same as asking how often they can avoid making an out every time up at the plate.

F-list

Loney 0%
Gonzalez 13%
LaRoche 14%
Pierre 15%
Kent 18%
Furcal 18%
Martin 20%
Abreu 21%
Garciaparra 22%
Ethier 26%
Kemp 26%
Betemit 32%

Loney has reached base at least once in every game in which he has at least two plate appearances. That sort of thing won't last, but it shows how good he's been so far. In this category Martin finally shows some weakness: he fails to reach base in 1 of 5 games. The reason Betemit has lost his third base job is his 1 in 3 games failing to reach base. Betemit is penalized severely by this grading method, though, since games with only one plate appearances don't count for anything. Kemp also finally shows some weakness, with almost a quarter of his limited games earning him an F.


These game grade distributions aren't a predictive tool at all. To answer the question of how a player is likely to do, you'll still want to start with the various batting percentages and also look at things like line drive rate and strikeout rate. These distributions are just another way, I hope a fun way, of seeing who has contributed offensively, and who hasn't.

Kent and Gonzalez are the only Dodgers to place in the top half of every list. These are the steady, mostly producing veterans. Martin, Kemp and Loney also all do very well, especially on the A and AB lists, showing again the promise and ability of these exciting young players. LaRoche finishes ahead of Abreu on all four lists, once again pointing out the absurdity of LaRoche being in AAA while Abreu plays on the big club. Do I even need to mention the showing of Pierre and Garciaparra? Nah. I'd rather look at Martin's numbers again.

5 comments:

Jon Weisman said...

Why don't you tally the scores like GPA: 4 points for an A, 3 points for a B, etc., and see what that gets you?

Joshua Worley said...

I'm guessing it gets a ranking very close to an OPS ranking! I thought about doing a GPA rank as well, but ... I didn't.

Here is is, with OPS also listed.

Loney 2.21 1.044
Martin 2.13 0.865
Kemp 2.13 0.902
Gonzalez 2.01 0.855
LaRoche 2.00 0.700
Kent 1.97 0.812
Furcal 1.83 0.695
Ethier 1.77 0.783
Abreu 1.71 0.721
Pierre 1.70 0.649
Betemit 1.62 0.802
Garciaparra 1.58 0.654


Hmm, it's not all that close to the OPS order. Don't know what it means, though, if anything. One point is does make again is that there isn't that much spread between the best and worst players ... until you start talking about a season's worth of games, I guess.

I'm just fascinated by how often players have good or bad games, since it is by games that we often evaluate players on a day-to-day basis. The thing that's missing is some context ... how do players on other teams do in this grading system? Another direction I'm looking at, if I ever do anything with this sort of thing again, is how Little and Colletti might form impressions of players based on the sorts of games they have. This might entail giving bonuses for RBIs, or discounting walks somewhat. Because obviously they feel Abreu typically has better games than LaRoche, which is completely contrary to my system. I guess they also value Abreu's versatility.

Jon Weisman said...

Yeah, I feel this is worth you continuing to pursue somehow.

Jon Weisman said...

Betemit's pinch-hitting heroism gets ignored using this method though, doesn't it?

Joshua Worley said...

Yes, Betemit is unfairly treated by the game grades. I mentioned that somewhere in the main post, though not as strongly as you put it.

I'm not one of those who wants to discount Betemit's pinch hitting exploits at all. It's just an unfortanate side-effect of this strange way of measuring offense that Betemit gets robbed.