Stupid Stats, Vol. 1

roubik2

I’m warning you: these will be stupid.

Over the last four years in MLB (2005-2008), the first hitter of the season for each team is hitting .306/.375/.454 (120 plate appearances)

Teams who get the first batter of the season on base (46 of 120) average 79.5 wins.

Teams whose first batter records an out (74 of 120) average 81.9 wins.

Denard Span led off the Twins 2009 season with a walk.  Teams that do that (12 of 120) average 80.7 wins. 4 of the 12 made the postseason.

Teams starting the season with a triple have not made the postseason in this time frame.  (the only example – 2006 Cubs – 66-96)

Every team whose first batter reaches on an error has made the postseason in this time frame (the only example – 2006 Dodgers – 88-74, NL West Champs)

7 of 29 teams whose season started with a strikeout made the postseason (which is pretty close to the percentage of teams overall – 7 of 32 = 21.9% while 29 of 120 = 24.2%)

The first hitter of the season for each team that eventually makes the postseason is hitting .286/.375/.464 (which mirrors almost exactly the overall numbers above)

In conclusion, none of this tells us anything, you just wasted all the time it took you to read this.

Sorry.

Playoff Wins Analysis

Given the data I posted over the last couple of days (Goals For, Goals Against, Power Play, and Penalty Kill vs. Playoff Wins) I figured I should take a look and see what trends emerge.

I divided things into quadrants, and looked at the top and bottom for each statistic.  With 48 playoff teams in the last three years, I was looking for about 12 teams in each sample.

Goals For –

Lowest Quadrant (< league average) – 14 teams, 5 series wins

Highest Quadrant (>10% above league average) – 12 teams, 18 series wins

Goals Against –

Lowest Quadrant (< league average) – 11 teams, 7 series wins

Highest Quadrant (>15% above league average) – 12 teams, 10 series wins

Power Play % –

Lowest Quadrant (< league average) – 14 teams, 7 series wins

Highest Quadrant (>10% above league average) – 11 teams, 19 series wins

Penalty Kill % –

Lowest Quadrant (<1% below league average) – 12 teams, 7 series wins

Highest Quadrant (>3% above league average) – 10 teams, 11 series wins

So that seems to contradict the saying “defense wins championships”.  The two categories that show the most difference betwixt the highest and lowest quadrants are Goals Scored and Power Play %.  So, you ask, what do those numbers look like on the same plot?

gf-ppX-axis is Goals Scored, and Y-axis is PP%.

So, teams that are less than 10% above average on both PP% and GF are not very likely to advance deep into the playoffs.  32 teams fit that description, and they have 17 series wins among them.  I don’t particularly like those odds (the outlier there is Edmonton’s trip to the Finals in ’06).  On the other hand, the teams more than 10% above average on both PP% and GF have 9 series wins among 7 teams. Leaving 19 series victories among the other 9 teams, so I guess you want your team to be really good at one of the two, but not both?  (Actually, that’s overanalyzing a really small sample size, so ignore that.)

So let’s look at the Western Conference of the current season and see who fits the mold of a possible playoff success story.

westHoly crap, Detroit is way out there (+29% GF, +42%(!) PP%).  Also Chicago and San Jose are >10% in both categories.  The rest of the conference is pretty mediocre, but a few teams have a good power play (St. Louis, Anaheim, and Minnesota are +13%, +12%, and +11% respectively in PP%).  Unfortunately, all three seem to lack the scoring punch in general (particularly the Wild) to make it to the playoffs, which is the first step toward a successful playoff run.

Now, I realize that predicting Detroit, San Jose, and Chicago as frontrunners in the conference doesn’t take this much analysis, but when I started making these plots I didn’t know exactly where it was going, and it was interesting to see how important offense is to a deep playoff run.

Finally, for those who are wondering, the East looks like this:

PP% >10% above average – Washington (+25%), Boston (+22%), Philadelphia (+21%)

GF >10% above average – Boston (+17%), Washington (+15%), Philadelphia (+12%)

Playoff Wins vs. Reg Season GF/GA

Data from the three post-lockout playoff runs.  Y-axis is playoff wins.  X-axis is percentage difference from average for that season.

gfGoals against is next (positive percentage means less goals allowed than average).

gaFinally, a plot of both GF and GA with color coding for advancement in the playoffs. (GF on X-axis, GA on Y-axis)

gf-gaTomorrow I’ll post the power-play and penalty kill numbers.  Hopefully, Thursday will bring some analysis of all this data as well as a look at the 2008-2009 season, and where current teams fit in this data.

Bullpen Usage

At the WGOM, ubelmann posited:

It would be interesting to chart something like winning percentage for home teams with an R run lead going into the 9th inning vs. season. I feel like I haven’t seen that anywhere? That would seem to be a strong indicator of whether or not modern bullpen usage is better than old-school bullpen usage.

That seemed like something I could jump on.

From 1977 to 2006, WE for th home team with a 1-run lead entering the 8th or 9th inning.

we-relieversNumbers from here.  Compare these numbers to this graph of bullpen usage to get an idea of when the philosophy changed concerning closers, and you get some surprising results.

1977 – 1989 (over 40% SV = 1+ IP)  –  9 of 14 seasons above average at preserving 1-run leads in the ninth. (2561 out of 2925 – 87.56% of leads preserved)

1990-2006 (over 40% SV = 1 IP)  –  4 of 16 seasons above average at preserving 1-run leads in the ninth. (3133 out of 3651 – 85.81% of leads preserved)

An artifact of a more offensive era, or a refutation of the current bullpen template of the closer as a “9th inning only” guy?

I lean toward the former, but it is pretty interesting that there’s a lack of a convincing argument for the modern bullpen.

Wild at the All-Star Break

Add one more thing to the list of things that mildly irritate me. Namely, when people compare the records of two teams that have played a different number of games. For most of the year the Wild have played about 2 or 3 games less than the Canucks, but everyone consistently places the Wild on the brink of the all important eighth position in the conference, well behind Vancouver, regardless of the fact that Minnesota has lost less games. So I generally keep track of losses (i.e. – missed points). That way if a team that is ahead in this regard wins the rest of their games, they finish ahead.

Right now the Northwest Division looks like this:

Calgary   = -32 pts (14 losses, 4 OT/SO losses)
Edmonton  = -41 pts (19 losses, 3 OT/SO losses)
Minnesota = -43 pts (20 losses, 3 OT/SO losses)
Vancouver = -45 pts (19 losses, 7 OT/SO losses)
Colorado  = -47 pts (23 losses, 1 OT/SO loss)

And the Western Conference like this:

San Jose    = -17 pts
Detroit     = -24 pts
Calgary     = -32 pts
Chicago     = -32 pts
Edmonton    = -41 pts
Minnesota   = -43 pts
Phoenix     = -43 pts
Dallas      = -43 pts
---------------------
Vancouver   = -45 pts
Columbus    = -45 pts
Anaheim     = -47 pts
Colorado    = -47 pts
Los Angeles = -47 pts
Nashville   = -49 pts
St. Louis   = -50 pts

That’s not a lot of space between 5th and last in the conference.

Looking back at the post-lockout standings at the midseason break, we get the following data.

2007-2008

(playoff teams in bold)


Detroit     = -24
San Jose    = -37
Minnesota   = -41
Calgary     = -42
Vancouver   = -43
Phoenix     = -44
Colorado    = -44
Dallas      = -45
-----------------
Nashville   = -45
St. Louis   = -45
Anaheim     = -46
Columbus    = -46
Chicago     = -50
Edmonton    = -53
Los Angeles = -60

2006-2007

(playoff teams in bold)

Nashville   = -27
San Jose    = -32
Anaheim     = -32
Detroit     = -33
Dallas      = -37
Calgary     = -38
Vancouver   = -40
Colorado    = -43
-----------------
Minnesota   = -44
Edmonton    = -46
Phoenix     = -50
St. Louis   = -50
Columbus    = -55
Chicago     = -55
Los Angeles = -62

2005-2006

(playoff teams in bold) – an Olympic year, so the break was later allowing for less shifting in the standings.

Detroit     = -31
Dallas      = -37
Calgary     = -41
Nashville   = -42
Vancouver   = -47
Colorado    = -48
Edmonton    = -48
Anaheim     = -49
-----------------
San Jose    = -50
Los Angeles = -51
Minnesota   = -55
Phoenix     = -60
Columbus    = -68
Chicago     = -70
St. Louis   = -71

Wow, 21 of 24 playoff teams in the last three years were at least tied for the eigth spot in the conference at the break.  The other 3 teams were a combined 3 points out of that 8 spot.  So it would seem to indicate that you had better have it figured out by the break, because we haven’t seen any team pour it on after to get in to the playoffs.

However, as tempting as it is to read too much into that and declare this year’s playoff teams set (after all, nobody is within one point of the eight seed, making those insurmountable leads, right?), I’m going to hold off on that for now.

Bullpen WPA

In the offseason, I think most Twins fans would agree that the bullpen is an area that the front office should be looking to improve. After Pat Neshek’s injury last season, the bullpen became a bit shorthanded, and everyone had to step up and pitch in situations that were a step up from what we had expected at the beginning of the year.

In an effort to grade how our relievers compared to others in the league, I looked at about 40 relievers, chosen by leverage index (i.e., how high-pressure were the situations they pitched in) and number of appearances. Thus I got a few comparison points for the six relievers used most often out of the Twins bullpen, these are all pitchers used in similar situations a similar amount of time.

To examine their effectiveness, I also tallied the number of appearances that resulted in a negative WPA. If they made it any more difficult for their team to win, I counted it as a negative outing. I also looked at “real bad” performances, which I defined as -0.80 WPA or lower. The -0.8 number is somewhat arbitrary, but 0.8 WPA is credited for closing a game pitching an entire inning with a 2-run lead. So I went with the negative of that value for symmetry’s sake.

Here’s the raw data, color coded to keep those used similarly together.

lev-index

To represent this graphically, everything was converted to percentage of negative WPA appearances.

negative

Also, a plot of the “real bad” performances as percentage of appearances with WPA < -0.8.

real-bad

Joe Nathan is awesome, Mariano Rivera is the only pitcher close to him in these representations.  Dennys Reyes had a lot of negative WPA appearances, although he was good at avoiding the blow-up outings (probably at least partly due to Gardenhire’s propensity to remove Reyes after a batter or two).  Guerrier and Crain both look pretty bad, while Bass is the middle of the road for relievers with his usage pattern (That usage pattern is pretty easy to replace, so not really the area we need outside help).  Breslow also had a pretty good year as far as preventing big, bad outings.

It is obvious that the loss of Neshek created some strain on the remaining members of the bullpen.  Coming into this year, however, Breslow had perhaps the best year last year of those analyzed here, and with Reyes gone, the Twins will need Breslow or Mijares or Crain/Guerrier to step up and improve the somewhat disappointing performance of last season.

Wild Help Me Avoid Public Embarrassment

Wild. It was a pretty good weekend for the Wild. After beating San Jose in overtime, they capitalized on that momentum and took the Red Wings to overtime, and the Red Wings even had to cheat (photo from Russo’s Rants) to push it that far. Most importantly, watching the Wild-Wings game with my friend from Michigan wasn’t nearly as embarrassing as it could have been, and it was downright fun for the first minute or so of the 3rd period when the Wild scored twice right out of the gate. Then a solid 2-0 win over Colorado on Sunday night. Owen Nolan had 3 of the Wild’s 4 goals this weekend, 2 of them power-play scores.

Some Owen Nolan PPG facts:

Since returning on New Year’s Eve Nolan has a PP goal in each of the three games.

One of every 6 power-play goals for the Wild comes from Nolan (6 of 36) and he has scored all 6 of his power-play goals in his last 9 games (granted he has missed time, only playing in 9 of the Wild’s 19 games in that span).

In the last 20 games, the Wild have 20 PPG. In the 10 games Nolan was inactive the Wild have 4 PPG and a 2-7-1 record. With Nolan active they have 16 PPG in 10 games (6 from Nolan himself) and a 6-3-1 record.

Another goal-scoring fact that doesn’t involve Owen Nolan at all:

Andrew Brunette has one goal in the last fourteen games with more than 15 minutes of ice time per game. Nonetheless he’s still second on the team in goals (ugh, the Wild can come up with some depressing statistics now and then).

Sunderland Black Cats. I started following them last year, but that’s a story for another time. They won their initial FA Cup match on Saturday over Bolton, a fellow Premier League team, so that felt pretty good. After last years early exits in both the Carling Cup and the FA Cup (one and done in both), it was good to see them reach the round of 16 in the Carling Cup and put a victory on the board in the FA Cup this year.

The draw for round 4 has taken place and the Cats get either Blackburn (from the EPL) or Blyth (from Conference North, the 6th tier of English soccer) who play each other on Monday (around 1 PM here in Chicago). As a Sunderland fan, I’m getting awfully tired of seeing Blackburn in these things, they were the opponent that knocked the Cats out of the Carling Cup this season. Couple that with two meetings already in the Premiership season (Sunderland took 4 of 6 possible points in those matches), and we could be seeing the true rubber match of this matchup soon.

Twins. More on them coming later this week. I’m reading Josh’s Top 50 Twins Prospects list as he puts them up. It’s some really interesting reading, and I’m looking forward to heading back to Beloit this summer for some Snappers games if I get the chance. This summer is going to be a little bit crazy though, so we’ll see how often I can get away.

Vikings. They just keep doing well enough that they won’t change their approach, but not well enough to go anywhere. I find myself caring less and less every year.