The recent AL Cy Young Award conversation focused attention on pitching's Triple Crown: ERA, wins, and strikeouts.  The latter remains one of the eye-catching measures of pitching performance, even though it’s generally held that it has limited effect on team success. 

That being the case, which team pitching staff has been the top strikeout (SO) staff in the AL? The argument will be made here that it is the Cleveland Indians staff of 1964-68. By what metric(s)? : 1) individual season SO records; 2) consecutive season SO records, analyzed to avoid against sampling error introduced by analyzing only too-few years; 3) SO records by comparison to the rest of the AL, which assure the leader is indeed striking out a greater proportion than the rest of the staffs during the same time period, not merely reflecting a high-SO period for all teams ; 4) SO records in comparison to the NL, which also emphasizes superiority not only to the AL, but to pitching across the major leagues; 5) Staff performance, not merely numbers pumped up by a one or two record-setting pitchers.

After 3 minor-league seasons, and 26 starts with the Tribe, Sudden Sam McDowell would join the club May 31, 1964, never to return to the minors.  Luis Tiant would make his first AL appearance July 19, replacing Tommy John on the roster.  Sonny Siebert would start his rookie season season in the bullpen, and work his way into the starting rotation by summer. The staff would SO 6.8 / 9 innings prior  to, and 7.2 / 9 innings after, June 1(1).This Big 3 would be the nucleus of the top SO staff for the entire 5 years. Gary Bell, Don McMahon, Pistol Pete Ramos, Jack Kralick, Dick Donovan, Ted Abernathy, and Floyd Weaver would all contribute to the SO totals.

The ’64 staff would be the firstAL staff to SO over 1100 batters, and over 7.0 / 9 innings. The Indians staff of ’64-68 set individual SO records that would last 30 years, until the steroid era.  Their yearly totals and SO / 9 innings are listed in Table 1.

                                          1964     1965      1966        1967      1968 

 Strikeouts                        1162*    1156      1111        1189       1157     

Strikeouts / 9 innings        7.03       7.13*      6.8           7.24*        7.1               

Table 1.  Yearly SO and SO / 9 innings of th Cleveland Indians Pitching staff 1964-68.

*AL Records

That staff would lead the rest of AL in SO by more than 20% in each of those years, and they did it in the ballpark with the largest foul area, Cleveland Municipal Stadium.  Only the California Angels would lead the AL more consecutive years.  But more on that in upcoming posts.

         1. STRIKE THREE! My Years in the ‘Pen, T.A.Tomsick, Cincinnati Book Publishers, Cincinnati, OH, 2010

<!--EndFragment-->

It is likely there will be two Wildcard playoff spots by 2013. To lambast the idea or to praise it as perfect is unwise. There are advantages and disadvantages to adding a play-in Wildcard game. However, the advantages do outweigh the disadvantages.

Continue reading this post »

Sorry for the long delay. College.

To refresh:

There is a new expansion draft format and for one day you get to be a GM of all 30 teams. Here's the catch; you can only protect three players from each organization (sorry not 15, then 3 more, etc., that would take way too long). You get the player as is with his current contract (it makes things more interesting). Keep in mind each teams financial statuses.

Today's version: AL Central

Much much more after the jump.

Continue reading this post »

Hey I am doing an economic research project on sabermetrics in division 3 baseball. The biggest problem that I am running into now is how to value a player in college baseball since they do not get paid. I can create most of the compound stats needed for sabermetrics but without a salary it is hard to decide if a player is undervalued, overvalued or correctly valued. Does anyone have any ideas as to how to give a value players in college?

I used my simulator to simulate every single game in the major leagues this season and am doing so for the World Series. I am posting the output of today's simulation, giving you a box score along with a table of the top 150 most likely final scores and the win expectancy of each team. My simulator incorporates a set of proprietary hitter and pitcher projections along with defense, park factors, weather, speed, splits, estimated actual lineups, a model for tiring pitchers and an engine for relieving pitchers based off of the score and the leverage index of the game (among many other things...). I simulated the game 100,000 times. You may notice that in the distribution of the likely final scores there are many one run home team victories near the top. This is due to the rules of baseball, with the home team batting last and sometimes winning by one run in walk off fashion. Also we run a simulation contest over at True Blue LA where you can wager pretend money on various prop bets from the game.  Feel free to join us.  Below is the data.

Simulation last ran Saturday at 8:00AM PST

Note: Lineups were estimated morning of game.

Box Score

Top 150 Most Likely Final Scores

Note: I have Lance Berkman and Mike Napoli playing DH in this simulation.

I used my simulator to simulate every single game in the major leagues this season and am doing so for the World Series. I am posting the output of today's simulation, giving you a box score along with a table of the top 150 most likely final scores and the win expectancy of each team. My simulator incorporates a set of proprietary hitter and pitcher projections along with defense, park factors, weather, speed, splits, estimated actual lineups, a model for tiring pitchers and an engine for relieving pitchers based off of the score and the leverage index of the game (among many other things...). I simulated the game 100,000 times. You may notice that in the distribution of the likely final scores there are many one run home team victories near the top. This is due to the rules of baseball, with the home team batting last and sometimes winning by one run in walk off fashion. Also we run a simulation contest over at True Blue LA where you can wager pretend money on various prop bets from the game.  Feel free to join us.  Below is the data.

Simulation last ran Wednesday at 11:00PM PST

Note: Lineups were estimated the night before.

Box Score

Top 150 Most Likely Final Scores

I used my simulator to simulate every single game in the major leagues this season and am doing so for the World Series. I am posting the output of today's simulation, giving you a box score along with a table of the top 150 most likely final scores and the win expectancy of each team. My simulator incorporates a set of proprietary hitter and pitcher projections along with defense, park factors, weather, speed, splits, estimated actual lineups, a model for tiring pitchers and an engine for relieving pitchers based off of the score and the leverage index of the game (among many other things...). I simulated the game 100,000 times. You may notice that in the distribution of the likely final scores there are many one run home team victories near the top. This is due to the rules of baseball, with the home team batting last and sometimes winning by one run in walk off fashion. Also we run a simulation contest over at True Blue LA where you can wager pretend money on various prop bets from the game.  Feel free to join us.  Below is the data.

Simulation last ran at 9:00PM PST

Box Score

Top 150 Most Likely Final Scores

To refresh:

There is a new expansion draft format and for one day you get to be a GM of all 30 teams. Here's the catch; you can only protect three players from each organization (sorry not 15, then 3 more, etc., that would take way too long). You get the player as is with his current contract (it makes things more interesting). Keep in mind each teams financial statuses.

Today's version: AL West

Much much more after the jump.

Continue reading this post »

Introduction

Now that the Major League Baseball season is over (and what a final day it was), we can start talking about who deserves what award. There is plenty of disagreement about who will actually win, whether or not pitchers should win the MVP and if candidates should come from a winning team. But regardless of all of that, who are at least legitimate candidates that we should be talking about?

Method

I like the Wins Above Replacement (WAR) framework, but each execution of it obviously has its flaws (evaluation of defense being the largest one). The two most well-known WAR statistics come from Fangraphs (fWAR) and Baseball-Reference (rWAR). So we can just average these two statistics and use that, right? Well, there still may be some error there. So I have decided to take the average of fWAR and rWAR and then create a +/- 15% error on this average. So Jose Bautista finished the season with 8.4 fWAR and 8.6 rWAR, for an average WAR of 8.5. Multiplying this number by .85 and 1.15 gives a High WAR of 9.8 and a Low WAR of 7.2. The true measurement of how well he played this season probably falls somewhere in there.

If we look at the top player's Low WAR and each of the next players' High WAR, we should find a good list of players who should be in the MVP or Cy Young conversation for their league. I don't believe that the vote should be based on WAR alone, but I do believe it is a good starting point.

Results

Continue reading this post »

Everyone knows that a pitcher’s win count is a terrible way to determine the skill of a pitcher. For example, you could have a pitcher on your team who pitches a complete game, strikeout 10, walked none, and give up one hit. But what if that one hit he allowed was a homerun? And what if your team has terrible offense, and didn’t score a single run in support of your pitcher? Well, then the pitcher gets the loss. Yup, even with his line of 9IP/0BB/10K/1H/1ER, he would still get the loss, because your team didn’t give him any run support. Likewise, you could have a pitcher who goes 5 innings, allows 10 hits, 5 walks, no strikeouts and, say, 10 earned runs. But as long as his team can score at least 11 runs, and the team’s bullpen holds on after his five inning outing, he will get the win. Doesn’t seem fair, does it? It really isn’t. Would you say the second pitcher is better just because he got the win? No, you wouldn’t, unless you are extremely stupid.

So, what’s the best way to determine the skill of a pitcher? Advanced ERA statistics such as FIP, xFIP, tERA, etc, seem to do a pretty good job. You can also check out a pitcher’s strikeout-to-walk ratio and opponent’s OBP. Another great stat is to use is WAR. If the pitcher has a WAR of roughly 5 or more for a season, chances are he had a good year. Likewise, if his season WAR is under 0, he was probably terrible, and his team would have been better using a pitcher from somewhere in their minor league system.

A real life example of using WAR to determine a pitcher’s skill is Nolan Ryan’s 1987 season against Jon Garland’s 2006 season, which is also further proof that pitcher wins are just terrible.

Nolan Ryan (1987): 8–16 win/loss; 2.76 ERA; 1.14 WHIP; 270 K; 87 BB; 5.5 WAR

Jon Garland (2006): 18–7 win/loss; 4.51 ERA; 1.36 WHIP; 112 K; 41 BB; 3.3 WAR

Looking at this, you can clearly see that Nolan Ryan was better, even though Jon Garland had a lot more wins and a lot more losses. But look at their WAR, Nolan Ryan was worth 5.5 wins above replacement (an All Star quality season), and Jon Garland was only worth 3.3 wins above replacement.

I’ll end this by saying there a truly a lot of ways to see how valuable a pitcher is, but please don’t use their win/loss record. It just isn’t fair. A pitcher can’t help if he’s on a team that can’t score runs. Leave a comment and tell me how you think the best way to determine a pitcher’s skill is. You can also find me on Twitter, @mitchell_jj. Thanks for reading.