Tuesday, August 31, 2010
Derek Jeter’s 2010 Actual vs. Projected Through August 30
Since Mel Hall mentioned this in an earlier thread and it was something I was planning to look at anyway, here’s how Derek Jeter’s 2010 year to date performance compares to his average projection heading into the season, pro-rated to his actual YTD playing time.
| Projection | PA | AB | H | 1B | 2B | 3B | HR | BB | HBP | SO | SB | CS | DP | AVG | OBP | SLG |
| average | 602 | 539 | 164 | 124 | 26 | 2 | 12 | 53 | 7 | 84 | 15 | 4 | 18 | .304 | .372 | .426 |
| ytd | 602 | 545 | 146 | 109 | 24 | 3 | 10 | 48 | 6 | 84 | 14 | 4 | 19 | .268 | .332 | .378 |
| difference | -6 | 18 | 15 | 2 | -1 | 2 | 5 | 1 | 0 | 1 | 0 | -1 |
Difference is just the projected stat minus the actual stat. So if you look at the hits column, Jeter has had 18 fewer hits than projected, 15 of them singles. He’s also hit two fewer HRs and doubles than projected, and walked five fewer times. K rate is spot on and SB/CS is right there. He’s hit into one more DP than projected.
If we think of Jeter’s average projection as a baseline, we can look at his projection in terms of a binomial distribution of each component stat.
| StDev | PA | AB | H | 1B | 2B | 3B | HR | BB | HBP | SO | SB | CS | DP | AVG | OBP | SLG |
| +2 Std | 602 | 539 | 189 | 129 | 36 | 5 | 19 | 68 | 12 | 66 | 22 | 0 | 10 | .351 | .447 | .541 |
| +1 Std | 602 | 539 | 176 | 126 | 31 | 3 | 15 | 61 | 10 | 75 | 18 | 2 | 14 | .327 | .410 | .483 |
| Baseline | 602 | 539 | 164 | 124 | 26 | 2 | 12 | 53 | 7 | 84 | 15 | 4 | 18 | .304 | .372 | .426 |
| -1 Std | 602 | 539 | 151 | 121 | 21 | 1 | 8 | 46 | 4 | 93 | 11 | 7 | 22 | .280 | .334 | .368 |
| -2 Std | 602 | 539 | 138 | 117 | 16 | 0 | 5 | 39 | 2 | 103 | 7 | 9 | 27 | .256 | .296 | .314 |
This is how I figure out my CAIRO percentile projections, although I also include a function to allocate playing time based on how a player’s projection looks (ie, a player who’s exceeding his projection will probably play more, a player who’s underperforming will play less). In this case since we’re looking at Jeter’s YTD performance, I’m holding the playing time constant.
As you can see, Jeter’s performance is within one standard deviation of his projection, albeit on the negative side.
So the question is probably whether or not those 18 missing hits and three missing extra base hits are the product of a decline in bat speed or just a representation of the fluctuation that can happen in a baseball season. Probably a bit of both, although perhaps the decline in bat speed is overstated somewhat by his actual results so far.
Comments
SG, fantastic work as usual. How did CAIRO nail the DP?
Can you get CAIRO to predict swinging strikes etc.? If he has more swinging strikes than predicted, that could be indicative of slower bat speed, no?
I’m surprised by the relatively small change in walks. What about BABIP, LD%, GB% if you have those?
SG, fantastic work as usual. How did CAIRO nail the DP?
Nothing fancy there, just a weighted average. All the systems that projected double plays were in the 17-19 range for Jeter.
Can you get CAIRO to predict swinging strikes etc.? If he has more swinging strikes than predicted, that could be indicative of slower bat speed, no?
As currently designed, no, but that could be something relatively straightforward to implement.
I want to look at Jeter’s PitchFX and see if we can get a more granular idea of what’s happening. A few things that I’ve pondered:
- I’m curious about how he’s hitting against above average fastballs. If his bat is slower, I’m assuming he’ll be missing them more and/or hitting more of them the other way.
- I’m also thinking that if he’s cheating (starting his swing early to compensate for a slower bat) it may mean he’s out in front on off-speed stuff. When you’re out in front you’re more likely to hit a ground ball.
- I also think a general comparison between this year and 2008/2009 might show some interesting things.
- I’m also wondering if he’s getting pitched inside more or struggling more against inside pitches.
- A general balls in play distribution might also show us if he’s not pulling the ball as much.
I’m pretty sure the PitchFX data would show us at least circumstantial evidence of a physical decline, but to do it right it’s probably going to take a bit of time to parse all the data. It’s on my to-do list though.
What about BABIP, LD%, GB% if you have those?
I don’t project those specifically, but a weighted average of 2006-2009 using the same weights I use for CAIRO gives us:
2010 Projected
BABIP: .361
LD%: 19.8%
GB%: 57.4%
FB%: 22.8%
2010 Actual
BABIP: .300
LD%: 16.8%
GB%: 65.6%
FB%: 17.5%
So Jeter has a higher GB% at the expense of FBs, which should generate a higher BABiP, but also a lower LD rate, which should generate lower BABiP.
The drop probably can be explained by weakly hit grounders. I do think his bat speed has slowed down.
Yeah, line drives are the most likely types of batted balls to be hits, something like 72% of the time. So a drop there is a bad thing.
Ground balls are hits something like 36% of the time and fly balls are hits 31% of the time, so in theory more GBs = more singles, although it also means fewer 2Bs and HRs.
Unfortunately, in Jeter’s case the ground balls sure seem to be hit with less authority than they have been in the past so it’s not working.
If Jeter would continue at his current pace I’m pretty sure he’ll end the season with more ground balls than any player has ever had in a season since it’s been tracked (2002).
Whatever the reasons for Jeter’s poor performance may be, this just goes to show that Kevin Costner was right. The difference between a good year and a crappy one is almost laughably small. One hit a week indeed.
What about expected BABIP?
Is our perception of 2009 vs 2010 warped by looking at WAR, assuming his measured defense was inflated by a big fluctuation? Is he worse this year on baserunning?
Do they actually measure bat speed or estimate declines based on production?
What about expected BABIP?
If you use the MLB average for hits on each batted ball type, Jeter’s BABIP should be around .319 instead of his current .300. Of course, if Jeter’s declined, we can’t necessarily assume he should be hitting for the MLB average on batted ball types.
A BABIP of .319 would give him 6 more hits, probably four singles and two doubles, which would move his line to .279/.342/.385. Still disappointing, but less so I guess.
Is our perception of 2009 vs 2010 warped by looking at WAR, assuming his measured defense was inflated by a big fluctuation?
WAR (at least Fangraphs’ version) is going to be heavily influenced by his defensive rating. On the individual player page you can look at the components that contribute to his combined WAR though.
2009
Batting: 37
Fielding: 6
Replacement: 24
Positional: 6
RAR: 74
WAR: 7.4
2010
Batting: -1
Fielding: -4
Replacement: 20
Positional: 5
RAR: 20
WAR: 2.0
Batting is just runs above an average hitter. Fielding is UZR. Replacement and Positional are the adjustments for replacement level and for playing SS. Those are counting stats so if he stays healthy Jeter should end up near the same in those two adjustments by year’s end.
So if you ignore the fielding and just look at the batting component, you can see that so far Jeter’s 38 runs worse than he was last year, although that gap should shrink some by season’s end.
Is he worse this year on baserunning?
According to Baseball Prospectus’s baserunning metric, he was worth about two runs above average last year and has been worth one this year, so not a big difference there.
Overall, Jeter has still been one of the top SS in the AL, even though he’s been so much worse than last year. That’s partly a testament to just how good he was in 2009, and partly a testament to a less than stellar crop of current AL SS.
He’s not really hurting the team. He’s just not helping them as much as he usually does, which means the Yankees need to get more contributions from other positions. Particularly if they’re going to commit to Jeter for the next few seasons.
I just checked Jeter’s hitting by month. He has gotten considerably worse in the latter part of the season.
2010 OPS by month
.875….....April
.702….....May
.718….....June
.621….....July
.662….....August
It’s worse in the second half, but it sucks pretty much all year, except April.
Question about the baseline: what was the offensive environment projected to be for that? According to the AIR stat on BBRef, Jeter’s environment has been a 98 this year, where 100 would be a .335/.400 (OBP/SLG). Last year was a 105. So a neutral baseline would make Jeter look a little worse, and using 2009 as the baseline would make Jeter look quite a bit worse.
Not that Jeter has played well. But maybe not quite as horribly as would be objectively expected. But then, when are we objective?
Question about the baseline: what was the offensive environment projected to be for that?
I think most projections just use last year’s run environment as the baseline, or a weighted average. Either one would be higher than what we’ve seen so far in 2010.
It’s true that Jeter’s line would probably look better if we account for the lower offensive environment this year, but I probably wouldn’t try to figure that until season’s end,
[14] Thanks.
SG, fantastic job as usual. Really brings home what a long season it is, and how little the differences that obsess us so… Re: #11 good job, and maybe Jeter is tired or has a small injury? And MC, I read you carefully this time. One hit a week…
Interesting and relevent history—Jeter Plight Conjures Up O’Neill/MattinglyMike Silva’s New York Baseball Digest
Do we know if Jeter is swinging at more pitches out of the zone this year?
Next entry: Athletics (65-65) @ Yankees (81-50), Tuesday, August 31, 2010, 7:05pm **Game Chatter**
Previous entry: A Tale of Two Long Relief Outings











