Looking Forward to 2008: The 2008 Playoffs
Last year, I reviewed the work of Vinay Kumar in an article at The Hardball Times web site
“So Billy, What Does Work in the Playoffs?” about how regular season statistics for a team could forecast its chances of success in the postseason. These showed that we were likely to get a Red Sox vs Cubs or Diamondbacks World Series, with the Red Sox winning it all. Well, I was half right (or two-thirds?). The Rockies challenged the statistical evidence in the NL Championship Series, but lost in the end.
So, building on last year's work, let's see how things have changed, and what that forecasts for the 2008 playoffs.
The Categories
Vinay used 30 categories in his original research. But rather than using the data straight up, he used minimum splits between two teams in order to eliminate about half of the results, to ensure that the data only reflected when a team had distinct advantage over its opponent. The columns below show the winning percentage in each category through 2006, and how the 2007 results changed them.
Team totals: through 2006 adding 2007
Won-lost record .581 .600
Runs Scored/Runs Allowed .537 .565
Batting records:
Runs scored total .415 .419
Batting average .447 .450
On-base percentage .452 .468
Slugging percentage .459 .452
Doubles .435 .453
Triples .442 .448
Home runs .476 .477
Batter walks .512 .543
Batter strikeouts (fewer) .587 .577
Stolen bases .512 .477
Stolen base attempts (more) .545 .522
Net stolen bases .378 .396
Stolen base percentage .306 .327
Caught stealing (fewer) .378 .396
Pitching records:
Runs allowed .605 .638
ERA .565 .592
Pitchers strikeouts .568 .551
Pitchers walks (fewer) .541 .550
Hits allowed (fewer) .732 .739
Home runs allowed (fewer) .588 .600
Complete games .628 .617
Pitchers shutouts .636 .667
Saves .457 .480
Saves by team leader .558 .574
Bullpen ERA .500 .545
Fielding records:
Errors committed (fewer) .643 .660
Defensive efficiency .659 .667
Fielding double plays .489 .500
Things never stay the same. Raw speed indicators continued to decline against Vinay's original research, although successful basepath larceny recovered a small amount of ground. Team totals continued their steady rise. Pitching recovered some of the value it lost in 2004-6. The ability to take or avoid a walk was a big factor last season. And bullpen depth and DER both continued to become more significant.
Last season I divided the categories into strong and weak ones, the strong ones being those where the teams holding the advantage have won more than half the playoff series.
Here are the strong categories:
Hits allowed
Defensive efficiency
Errors committed (fewer)
Pitchers’ shutouts
Runs allowed
Complete Games
Won-lost record
Home runs allowed (fewer)
ERA
Batter strikeouts (fewer)
Saves by team leader
Runs Scored/Runs allowed
Pitcher strikeouts
Pitcher walks (fewer)
Bullpen ERA
Batters' walks
There is an extra category this year. Stolen base attempts has fallen out of the strong categories, to be replaced by Bullpen ERA and Batters' walks, which are almost equal on percentage of teams holding a significant advantage going on to win their series.
Let’s carry this information forward and profile the 2008 Divisional Series. Before I take the prognostication plunge, however, a caveat – this year's data is nowhere near as clear cut as last year's. In two series we're going to see a key test of the principle that it is better to have an advantage in strong categories (which will appear in bold) than to be disadvantaged there, but have the advantage once both strong and weak categories are assessed. So bear that in mind as you read.
Milwaukee vs. Philadelphia
These two teams are evenly matched, with few decisive advantages, strong or weak. However, looking at this year's data I began to wonder about the influence of park factors. The Phillies have one of the top four pitching staffs in the National League, yet we may still think of Citizens' Bank Park as Coors East. (Although it's more middle of the pack this year, on single season park factors.)
Brewers' advantages
Doubles
Complete Games
Double Plays
Philadelphia advantages
Batter's Strikeouts
Net Stolen Bases
Stolen Base percentage
Caught Stealing
Saves by team leader
Errors committed
PREDICTOR PICK: PHILADELPHIA.
Hedging my bet: The running game is a recipe for playoff failure.
Chicago Cubs vs. Los Angeles Dodgers
This series is remarkable for the sheer quantity of advantages. No fewer than twenty-three come into effect here, and it's tightly balanced. What this might tell us is that both teams have serious weaknesses, and the winner is going to be the team that exploits their opponents' shortcomings most effectively. That is, this is a series to be lost rather than to be won. It's Rome vs Parthia, the Crusaders vs Saracens, as two armies from different military backgrounds clash. Last season the Cubs were in a similarly tough series to predict, and I went with them because although they were worse in the overall score, the teams were level on strong categories and the Cubs' categories generally had higher winning percentages. This year, I'm sticking strictly by the numbers.
Cubs’ advantages
Won-lost record
Runs scored/runs allowed
Runs scored
Batting average
On-base percentage
Slugging
Doubles
Batter walks
Saves
Saves by team leader
Defensive efficiency
Los Angeles Dodgers' advantages
Triples
Batter strikeouts (fewer)
Stolen bases
Stolen base attempts
Net stolen bases
Pitcher walks
Home runs allowed (fewer)
Complete games
Shutouts
Saves
Bullpen ERA
Double plays
PREDICTOR PICK: LOS ANGELES DODGERS
Hedging my bet: This year, maybe controlling top categories like Defensive Efficiency will work to the Cubs' advantage.
Boston vs Los Angeles Angels of Anaheim
Last year, Boston just dominated the Angels across the board. This season, things are a lot closer. The Angels are still behind in the overall total, but they've chosen their advantages with care. However, Boston controls the best overall category, Hits Allowed, and the fashionable Batter Walks. Nonetheless, this Boston team resembles those Yankee teams that have crashed out of the playoffs short of the World Series, relying on hitting categories for an overall advantage, but not controlling that many key categories.
Boston advantages
Runs scored
Batting average
On-base percentage
Slugging percentage
Doubles
Triples
Home Runs
Batter walks
Hits allowed
Pitcher shutouts
Angels’ advantages
Won-lost record
Batter strike outs
Pitcher walks
Saves
Saves by team leader
PREDICTOR PICK: LOS ANGELES ANGELS of ANAHEIM
Hedging my bets: Boston keeps the hits down, and the Angels can't take the walk.
Chicago White Sox?Minnesota Twins vs. Tampa Bay
The Rays are good. They are right up there with the Red Sox on offense, and their pitching is pretty good in some key categories. They dominate the Twins/White Sox this year like Boston dominated just about everybody last year. The White Sox, meanwhile, can only look to avoid striking out while keeping walkers off the basepaths, as they wait around for a dinger to settle things – rather like the tiebreaker with the Twins.
Tampa Bay advantages
Won-Lost record
Triples
Batter Walks
Stolen bases
Stolen base attempts
Net stolen bases
Stolen base percentage
Runs allowed
Hits allowed
Complete games
Shutouts
Saves
Bullpen ERA
Errors
Defensive Efficiency Rating
Strikeouts
Chicago White Sox' advantages
Home Runs
Batter Strikeouts
Caught stealing
Pitcher walks
PREDICTOR PICK: TAMPA BAY RAYS
Hedging my bets: Detroit had a 9-3 advantage over St. Louis in 2006, so anything can happen.
Peering Further Ahead
Do you remember the 2005 ALCS? Can you argue that the Angels lost on a bad umpiring call? If you answer those questions 'yes', then you're in the mood for this postseason. Past the Divisional Series, things are tight. Following the logic of my numbers above, we'll get two extremely balanced Championship Series, as tight as that 2005 ALCS was on paper. (It was 2-1 to the Angels on the strong categories, and 8-4 to them overall. But the strong indicators are better at dealing with teams' relative strengths.) Tampa Bay and Anaheim are but one strong category apart. Philadelphia and Los Angeles are equal.
However, Philadelphia have a big advantage overall, and I'm inclined to go with that. Mostly, I'm surprised how good their pitching is. In the AL, I think Tampa Bay's offense gives it the edge. And who wins a Tampa Bay vs Philadelphia series?
Tampa Bay. 2008 - year of the Ray.
fra paolo
Posted: September 30, 2008 at 11:04 PM |
14 comment(s)
Related News:
General
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
I'm curious why you conclude this. The numbers above give a solid lean to the team that tries to steal more. It would seem to be that attempted SB would be the best stat to determine a team's commitment to the running game. The other SB stats are low, but not as relevant to a team's "running game" as attempted SB.
Great stuff, and (one time only)...Go Phils!
I'm not sure fra paolo does actually conclude that. These series look very close compared to other years and that was his bet-hedging comment while his predictor pick went the other way.
Stolen base attempts and Stolen bases used to be more decisive factors in years gone by but their advantage seems to be decreasing - quite sharply in fact with 2007's data. Last year at least, they were terrible categories to have the advantage in - it looks like they were strong indicators that the team would lose.
He didn't conclude it so much as state it as a fact. I, too, wonder why that should be so (although I am certainly prepared to believe it, why not).
All the baserunning indicators are below .500 except Stolen base attempts and even that fell out of the strong categories (as successful stealing goes in the opposite direction). The only reason that I can think of why Stolen base attempts should be a positive indicator (while the others are negative) is that potential basestealers mess up the defense and that is more significant than the actual success of the stolen base. This could be linked to the data relating to teams' likelihood of committing errors.
Teams that rely significantly more than their opponents on the running game to win baseball matches have come unstuck more often than not in the playoffs. Why that should be could be a number of factors - playoff pitchers are better at holding runners, backup catchers don't take much part in the playoffs, managers are less likely to take a chance, or the defending managers are more aware of the dangers of conceding a base. It would be interesting to see the data from the point of view of the defending teams.
Well, ok, and awkward phrase....but, it's a statement based (presumably) on the information presented, so that sounds like a conclusion to me.
If it's not based on the data (i.e., a conclusion) then I'm just curious why in general one would thing the running game is a recipe for disaster. There seem to be many other ways to hedge one's bets, specifically, believing that the Phils' bullpen has been extraordinarily lucky this year.
Eddings and Pierzalksljasfjsadfski are still ##########, though.
IOW, a good 5th and 4th SP will get you into the playoffs, but doesn't work in the playoffs.
Since the other three division series are nearly all decided (Rays 4 up with 2 innings to play) and paolo's predictor is (almost) 3 for 3, let's see how the fourth one goes.
Of course, the weighting will change after these four series. We may even see the formula change the author's predictions for the championship series. It will be interesting to see how the new predictor picks look.
Padraic wanted to know why I concluded the running game didn't work in the playoffs. I almost certainly exaggerated the point I wanted to make, but the trend since 2003 has been a loss of effectiveness in teams for which the running game is an important weapon. Vinay's numbers were a lot higher than the current ones, and coming after the victory of the Marlins, it suggested that the Holy Trinity of the non-sabermetric - pitching, speed and defence - was the stuff that worked in the playoffs. This is less true now, and the trend until this season had been steadily down. Hence my comment. Buying into speed was buying into a falling market.
For John DiFool, I'd say I'm not sure I'd buy the methodology either. (Although I have used it to make bets, so maybe I have.) One could probably throw out a few categories because they tend to appear together. Caught Stealing (fewer) and Net Stolen Bases have been identical for awhile, but they've now diverged. Some categories drift up and down, but not decisively in either direction. All I can say is that this survey has worked so far, especially in the Division Series, but there are failures. Which brings me to:
I thought Boston's domination in the most decisive categories, hits allowed and pitcher shutouts, would be more influential...
philistine did something this year that I did last year when I backed the Cubs over the Diamondbacks with exactly this kind of reasoning. I also might well have backed the Mets over the Cardinals in 2006, as the Mets had a massive lead, and held the two strongest categories, hits allowed and pitcher shutouts. Well, we both would have lost then. My empirical observation is that the quality of categories is not as significant as the quantity. But sometimes neither works.
NL Championship Series
Philadelphia's massive lead overall means they should win this, although I think they have to win both opening games to do so. In spite of what the statistics say, the Phillies pitching might be better than the Dodgers', who have a big advantage in their home park. Yet the Dodgers team that compiled the full-season statistics and the Dodgers team in this postseason are two different creatures, and I believe have a good shot at proving the Predictor wrong. So I imagine if they can take one in Citizens Bank Park, they'll be able to nullify the strong Phillies offense at Dodger Stadium sufficiently to take the series. Of course, the Predictor has to go with the numbers.
Philadelphia (13-4)
Won-lost record
Runs scored/against ratio
Runs scored
Slugging
Doubles
Triples
Home Runs
Net stolen bases
Stolen base percentage
Caught stealing (fewer)
Saves
Saves by closer
Errors
Los Angeles Dodgers (4-4)
Batter strikeouts (fewer)
Pitcher strikeouts (more)
Pitcher walks (fewer)
Home runs allowed (fewer)
Predictor Pick: Philadelphia Phillies
AL Championship Series
Well, to cut to the chase, the Red Sox burst the Rays' bubble. I just don't like the look of the Rays' advantages against the Red Sox. They don't control any hitting category, and I only see them winning if they can take an early lead of two or three runs, in which case I think they could defend it. But how to do that without a hitting advantage?
Boston (8-3)
Runs scored
Batting average
On base percentage
Doubles
Batter strikeouts (fewer)
Caught Stealing (fewer)
Pitcher shutouts
Saves by closer
Tampa Bay (5-2)
Stolen bases
Stolen base attempts
Saves
Bullpen ERA
Defensive Efficiency Rating
Predictor Pick: Boston Red Sox
And who comes out on top in a Phillies v Red Sox series?
Red Sox repeat.
Interesting change to WS prediction now, with a rock, scissors, paper scenario. Angels should beat Sox who should beat Rays who should beat Angels.
In the other prediction thread, I went for Dodgers over Phillies as a hunch and followed fra paolo's earlier predictor pick for the AL - Rays. So now I'm torn between wanting to see the predictor succeed and hoping my own picks turn out right.
You must be Registered and Logged In to post comments.
<< Back to main