Statistical analysis of wins and losses

[Balta1701]

Well, we're now 71 games into the 2005 season, the season that was supposed to bring a new style of baseball to the South Side, and a season that has certainly been a fun one to watch.

 

Now that we're nearly half way into the season, and since today's an off day, I think it's as good a day as any to take a look at how this team actually has changed from last year's team in the most important categories around; how many runs we put on the board compared to the amount of runs the other team puts up.

 

For this analysis, I used results from ESPN.com for my 2005 results and Baseball Almanac.com for my 2004 results.

 

My work was as follows: I entered the final score for each game over the last 2 years into Excel and performed some rudimentary statistical analyses on the scores to see if any obvious trends appeared. And, not to my surprise, I found several.

 

First, let's look at the average amount of runs scored:

 

2004: 5.339506173

2005: 4.845070423

 

Ok, so for starters we're down about a half a run per game. That's not surprising to me at all, considering Lee is in Milwaukee, Frank's played like 10 games, and Ordonez played half a season last year.

 

But let's look past total runs scored for a minute, and look at 2 other numbers:

 

Median runs scored:

2004: 4.5

2005: 5

 

Standard Deviation of Runs Scored:

2004: 3.721956976

2005: 2.707913732

 

A-Ha! Here I present to you what I feel the key stats are this year for our offense. If you look at the median runs scored, in 2004 it is right between 4 and 5, meaning 1/2 of our games we scored 5 runs or more, and 1/2 we scored less. In 2005 on the other hand, we're scoring 5 runs or more a fair amount more often, and we're scoring 4 runs or less in fewer games. This is good; if you imagine our pitchers averaging 5 runs allowed per game like they did last year, the more games we score over 5 the more we're going to win.

 

The second stat is the real key to this offense; this is the "Speed doesn't go into slumps" stat, this is the Ozzieball stat.

 

Last year, we scored 5.3 plus or minus 3.7 runs per game, on average...with a few games falling outside those bounds.

 

This year, we're scoring 4.8 runs a game, plus or minus 2.7. We've decreased our standard deviation by a full run. This means that yes, we're scoring less 15 run games, but it also means we're having less 0 and 1 run games. The numbers are clustering much more closely around 4, 5, and 6 runs per game; numbers that give us a good shot to win the game.

 

Looking at these numbers...it is clear to me that Ozzieball is actually working; we're scoring enough runs to win more often, even though we're not scoring as many total runs as we were last year. The variation in our runs scored has dropped, and we're scoring 5 runs or so a lot more often than last year's ballclub.

 

1 last item I noticed...and I can't figure this one out at all. For some reason, the White Sox in both 2004 and 2005 just do not like to score 3 runs. In 2005 we have scored 3 runs exactly 4 times, while we've scored 2 or 4 runs 23 times between them. The same trend happened in 2004: we scored 2 runs 24 times, 4 runs 24 times, but 3 runs 12 times. It's a really wierd pattern in a histogram, and I have no idea why it's happening. I also looked at our opponents scores, and they are much closer to a bell curve, while the Sox scores have these 2 peaks with a large dip at 3 runs per game.

 

So, next time the Sox score 3 runs in a game...act surprised. It really is a rare occurence.

Edited June 23, 2005 by Balta1701

[Gene Honda Civic]

[wsox08]

Nice work guys. I couldn't have done it better myself

[3E8]

First, let's look at the average amount of runs scored:

2004: 4.845070423

2005: 5.339506173

Are these years flip-flopped?

[Balta1701]

Are these years flip-flopped?

 

This is what happens when you copy & paste from Excel. Fixed!

[DABearSoX]

I wish I would have paid attention in Statistics.....

[danman31]

That really does explain it best. Nice use of stats while keeping it simple.

[Jabroni]

Buster Olney of ESPN just said that the White Sox will win the A.L. Central because of some statistic that most teams win with a record as good as ours at this point in the season. It was something like that.

Edited June 23, 2005 by Jabroni

[danman31]

Buster Olney of ESPN just said that the White Sox will win the A.L. Central because of some statistic that most teams win with a record as good as ours at this point in the season.  It was something like that.

No team that has ever had a 10 game lead before the all star break has failed to at least win their division or league back when that applied.

[southsider2k5]

Great work Balta. Nice thread.

[daa84]

Well, we're now 71 games into the 2005 season, the season that was supposed to bring a new style of baseball to the South Side, and a season that has certainly been a fun one to watch.

 

Now that we're nearly half way into the season, and since today's an off day, I think it's as good a day as any to take a look at how this team actually has changed from last year's team in the most important categories around; how many runs we put on the board compared to the amount of runs the other team puts up.

 

For this analysis, I used results from ESPN.com for my 2005 results and Baseball Almanac.com for my 2004 results.

 

My work was as follows:  I entered the final score for each game over the last 2 years into Excel and performed some rudimentary statistical analyses on the scores to see if any obvious trends appeared.  And, not to my surprise, I found several.

 

First, let's look at the average amount of runs scored:

 

2004:  5.339506173

2005:  4.845070423

 

Ok, so for starters we're down about a half a run per game.  That's not surprising to me at all, considering Lee is in Milwaukee, Frank's played like 10 games, and Ordonez played half a season last year.

 

But let's look past total runs scored for a minute, and look at 2 other numbers:

 

Median runs scored:

2004:  4.5

2005:  5

 

Standard Deviation of Runs Scored:

2004:  3.721956976

2005:  2.707913732

 

A-Ha!  Here I present to you what I feel the key stats are this year for our offense.  If you look at the median runs scored, in 2004 it is right between 4 and 5, meaning 1/2 of our games we scored 5 runs or more, and 1/2 we scored less.  In 2005 on the other hand, we're scoring 5 runs or more a fair amount more often, and we're scoring 4 runs or less in fewer games.  This is good; if you imagine our pitchers averaging 5 runs allowed per game like they did last year, the more games we score over 5 the more we're going to win.

 

The second stat is the real key to this offense; this is the "Speed doesn't go into slumps" stat, this is the Ozzieball stat.

 

Last year, we scored 5.3 plus or minus 3.7 runs per game, on average...with a few games falling outside those bounds.

 

This year, we're scoring 4.8 runs a game, plus or minus 2.7.  We've decreased our standard deviation by a full run.  This means that yes, we're scoring less 15 run games, but it also means we're having less 0 and 1 run games.  The numbers are clustering much more closely around 4, 5, and 6 runs per game; numbers that give us a good shot to win the game.

 

Looking at these numbers...it is clear to me that Ozzieball is actually working; we're scoring enough runs to win more often, even though we're not scoring as many total runs as we were last year.  The variation in our runs scored has dropped, and we're scoring 5 runs or so a lot more often than last year's ballclub.

 

1 last item I noticed...and I can't figure this one out at all.  For some reason, the White Sox in both 2004 and 2005 just do not like to score 3 runs.  In 2005 we have scored 3 runs exactly 4 times, while we've scored 2 or 4 runs 23 times between them.  The same trend happened in 2004:  we scored 2 runs 24 times, 4 runs 24 times, but 3 runs 12 times.  It's a really wierd pattern in a histogram, and I have no idea why it's happening.  I also looked at our opponents scores, and they are much closer to a bell curve, while the Sox scores have these 2 peaks with a large dip at 3 runs per game.

 

So, next time the Sox score 3 runs in a game...act surprised.  It really is a rare occurence.

good work, especially finding the median. thats important imo. i also think the standard deviation doesnt tell all that much. when taken, the deviations starting point is the average, but you can only go down 5.3 runs, where as on the upside, you can go up as many runs as the team had scored 18 i believe was last years high. therefore if you have a team that scores a bunch of runs alot, like the white sox had, the standard deviation will be a larger number, due to the greater absolute value of runs on the high side of the average as opposed to the lower side of it. therefore it really only emphasizes the feast aspect of the 2004 chisox, and fails to capture the entire famine portion of that offense as well. however, the work is well done and we can still take away the fact that the 2004 white sox were far less consistent in scoring runs than the 2005. i also saw something, i wanna say by dayn perry, but i could be wrong, taht showed that the white sox have scored 0 - 2 runs fewer than any other team in the majors. and on top of that their winning % in those games is astounding taken in consideration to the average w% of such games

Edited June 24, 2005 by daa84

[JUGGERNAUT]

IMO, the general statistics don't tell you much about the success of this team. The two key statistics are RON, & CL&L. I never paid attention to them as much as this year. They might just be the most important team stats in MLB.

 

RON 05 CWS 281RBI .347O .459S vs Opp 209RBI .309O .360S : +72 RBI

CLL 05 CWS 062RBI .349O .431S vs Opp 052RBI .238A .317O : +10 RBI

 

RON 04 CWS 684RBI .348O .471S vs Opp 667RBI .351O .460S : +17 RBI

CLL 04 CWS 119RBI .339O .445S vs Opp 112RBI .338O .446S : +07 RBI

 

Compare this with last yr's Twins:

RON 04 MIN 618RBI .339O .430S vs Opp 583RBI .333O .414S : +35 RBI

CLL 04 MIN 125RBI .340O .397S vs Opp 119RBI .325O.327S : +06 RBI

 

Our opponent in the WS this year?

RON 05 STL 309RBI .357O .439S vs Opp 235RBI .331O .382S : +74 RBI

CLL 05 STL 042RBI .320O .358S vs Opp 027RBI .285O .301S : +15 RBI

 

We lead STL by 4.5 gms for the best record in MLB & yet they hold the edge in RBI's in both RON & CL&L. They might be better than us but we've got lady luck on our side.

 

No comparison would be complete today w/out the Cub:

RON CUB 05 270RBI .343O .469S vs Opp 254RBI .348O .413S : +16 RBI

CLL CUB 05 048RBI .333O .426S vs Opp 049RBI .344O .380S : -1 RBI

Contenders? Who are they kidding?

[danman31]

You know it might be helpful to tell us what the f*** RON and CL&L are. Runners on and close and late?

Edited June 24, 2005 by danman31

[JUGGERNAUT]

You know it might be helpful to tell us what the f*** RON and CL&L are. Runners on and close and late?

<{POST_SNAPBACK}>

 

Yes.