NLD XXXVI - The Fun Continues

Desmond had a good series v AT so I have my fingers crossed that that will continue.

Meanwhile Wainwright (5.17 ERA) continues to tinker with his mechanics. We're hoping he can get in a good groove and earn his big contract money.

The league is past the quarter pole and the divisions are pretty close, The Central is amazingly close. The other two divisions are shaping up as two team races, but the other teams are only one hot streak from being right there.

I did some more thinking about the relationship between "runnable hits" ("RH"), and baserunning opportunities. For this post, I'm arbitrarily calling all singles and doubles "runnable hits", even if the game system causes some to be opportunities, and some not, randomly or otherwise.

I took the team stats (offense) after last night's games (game 45), and did some tweaking to look for ratios. This is what I found (in order of most runnable hits to least):

Team // Total singles and doubles // Baserunning Opp's (per strat stats) // Percent of opp's to RH's // Variance to avg.// Percent variance to average

The overall average of baserunning opportunities to runnable hits was 24.1%, for all teams.

Winnipeg // 386 // 90 // 23.3% // -.8% // -3.3%
Deep // 346 // 78 // 22.5% // -1.5% // -6.9%
Naples // 345 // 87 // 25.2% // +1.1% // +4.5%
Philly // 337 // 89 // 26.4% // +2.3 % // +8.8%
Mo' Bay // 331 // 78 // 23.6% // -.5% // -2.2%
Ottawa // 330 // 81 // 24.5% // +.5% // +1.9%
OCity // 308 // 79 // 25.6% // +1.6% // + 6.1%
Hudson Valley // 306 // 79 // 25.8% // +1.7% // +6.7%
War // 302 // 72 // 23.8% // -.2% // -1.0%
Burgh // 300 // 61 // 20.3% // -3.8% // -18.5%
Forest City // 291 // 71 // 24.4% // +.3% // +1.3%
Bonn // 291 //68 //23.4% // -.7% // -3.1%

This is only analysing baserunning opportunities. It doesn't analyse how often runners try to advance, or their success rate.

The only team that is really outside the normal ratio here is Burgh. Not sure why, but there it is. Otherwise, every other team is fairly close to the average ratio of 24.1%.

What this means is that, according to strat's statistics, baserunning opportunities arise on average about 24% of the time that singles or doubles occur. This would reflect the percentage of time that any runner is already on first or second base when the actual base hit occurs. It would also reflect the percentage of times that the base hit actually results in a baserunning opportunity in the game system. As Joe suggested, some SINGLE** results still seem to be automatic 2 base advancement, despite the "max" rules used in the online game. I haven't tried to analyse it yet, but maybe some DOUBLE** results also remain only 2 base advancement, with no option to take the extra base - maybe.

I think it's interesting that there seems to be some kind of a "normal relationship" between singles and doubles being hit, and baserunning opportunities. It's also interesting that the ratio is about 24% of the time. This ratio might be helpful in trying to come up with a mathematical theory about the run production value of baserunning and outfielders' arms.

But, GB, to come up with a mathematical theory, you need to know what percentage that the arm comes into play, period. So you have to know how often that a 1b or 2b happens. Then, you have differentiate between the three OF positions because they don't happen equally and, as you pointed out, there is a +2/-2 difference in throwing to 3b from RF or LF.

But, GB, to come up with a mathematical theory, you need to know what percentage that the arm comes into play, period. So you have to know how often that a 1b or 2b happens. Then, you have differentiate between the three OF positions because they don't happen equally and, as you pointed out, there is a +2/-2 difference in throwing to 3b from RF or LF.

If it IS random, then I really despair of ever learning the precise degree of randomness. The problem is that it would be difficult to learn that percentage by inference. To do that, you would have to know precisely what % cut-off strat uses for base running managerial decisions, and whether it is influenced by the game score, and then how it is influenced by the game score. Then you would have to analyze game results and determine what percent of the time strat's base running game results suggests that the runner COULDN'T advance, despite the max rules, by random reduction. In short, I think it's practically impossible for the average strat fan - on the outside of their game system looking in - to try to make an inference from analyzing specific game results.

The next best thing to knowing the actual rate of occurrence of the random event, IMO, is to know the rate at which base running opportunities occur in the total population of base running potential events. That is, the global rate of occurrence, without trying to analyze it by specific situational game results. That's what I'm trying to do with that analysis. If there is a relatively constant rate of occurrence of base running opportunities over total theoretical opportunities, then I can use that to construct a base running run production theory, I think. That's my idea now, anyway. And if there is a constant (say 24%) - practically speaking anyway, even if it isn't programmed that way - then why not use it as part of the model?

When I get discouraged with one approach, I'll try a different one!

But, GB, to come up with a mathematical theory, you need to know what percentage that the arm comes into play, period. So you have to know how often that a 1b or 2b happens. Then, you have differentiate between the three OF positions because they don't happen equally and, as you pointed out, there is a +2/-2 difference in throwing to 3b from RF or LF.

If it IS random, then I really despair of ever learning the precise degree of randomness. The problem is that it would be difficult to learn that percentage by inference. To do that, you would have to know precisely what % cut-off strat uses for base running managerial decisions, and whether it is influenced by the game score, and then how it is influenced by the game score. Then you would have to analyze game results and determine what percent of the time strat's base running game results suggests that the runner COULDN'T advance, despite the max rules, by random reduction. In short, I think it's practically impossible for the average strat fan - on the outside of their game system looking in - to try to make an inference from analyzing specific game results.

The next best thing to knowing the actual rate of occurrence of the random event, IMO, is to know the rate at which base running opportunities occur in the total population of base running potential events. That is, the global rate of occurrence, without trying to analyze it by specific situational game results. That's what I'm trying to do with that analysis. If there is a relatively constant rate of occurrence of base running opportunities over total theoretical opportunities, then I can use that to construct a base running run production theory, I think. That's my idea now, anyway. And if there is a constant (say 24%) - practically speaking anyway, even if it isn't programmed that way - then why not use it as part of the model?

When I get discouraged with one approach, I'll try a different one!

On the other hand, I could

(a) try to nail down - with strat's input - what items are included (or not included) in the base running opportunity stats (eg. XChart "decide" results?/ FB(b)? results? etc), and then I could:

(b) continue what I did before to calculate total theoretical opportunities. Then I could:

(c) compare the stats for opportunities to theoretical opportunities, to get the random reduction percentage.

That would work - assuming I can get solid input from strat on what's included (or not included) in their stats for base running opportunities.

I would suggest that you could assume that the random percentage approximates the actual percentage of chances on the real cards (adding and subtracting). Maybe that helps. It *would* be interesting to know what SOM actually includes and what the random chance it.

The OTHER thing that would be very interesting is to know exactly how the home field advantage works. Everything I've read is very generic and hard to figure into any kind of calcs.

I would suggest that you could assume that the random percentage approximates the actual percentage of chances on the real cards (adding and subtracting). Maybe that helps. It *would* be interesting to know what SOM actually includes and what the random chance it.

The OTHER thing that would be very interesting is to know exactly how the home field advantage works. Everything I've read is very generic and hard to figure into any kind of calcs.

I agree on all points!

The homefield advantage is particularly weird.

Joe, you said in one of your posts that you used to play lots of face to face games. So did I! That's the true genius of this game - it was (and still is, mostly) completely understandable! "Almost" every aspect can be explained in very precise terms. These "mystery" results (which lurk somehow in some influence of the die rolls - arggh) are a bit disturbing to people like you and me, who grew up playing the game with actual dice!

Anyway, I hope to get some more guidance. We'll see!

I would suggest that you could assume that the random percentage approximates the actual percentage of chances on the real cards (adding and subtracting). Maybe that helps. It *would* be interesting to know what SOM actually includes and what the random chance it.

The OTHER thing that would be very interesting is to know exactly how the home field advantage works. Everything I've read is very generic and hard to figure into any kind of calcs.

I agree on all points!

The homefield advantage is particularly weird.

Joe, you said in one of your posts that you used to play lots of face to face games. So did I! That's the true genius of this game - it was (and still is, mostly) completely understandable! "Almost" every aspect can be explained in very precise terms. These "mystery" results (which lurk somehow in some influence of the die rolls - arggh) are a bit disturbing to people like you and me, who grew up playing the game with actual dice!

Anyway, I hope to get some more guidance. We'll see!

Hey GB, I'm still playing in a FTF league! It's still fun to roll the dice and see what happens, but sometimes, I think the computer is more fair than *my* 20-sided die rolling!

Let us know what you find out!

I sent to strat-o-matic a very detailed analysis of my research on baserunning, and on the statistical data published by strat on the team "misc" page.

I'll publish it here as well, for your interest (I'm assuming that, I guess!).

Joe (and anyone else), if you have any thoughts or critiques on this, I'm "all ears".

Here it is:

This is my follow-up analysis on what appears to be a systemic error on the Strat statistics for team baserunning, on the Misc page for the team. Alternatively, the discrepancy might reflect some randomized application of the max rules, or some other application of the max rules. Could strat-o-matic ("strat") please provide me with an explanation of the discrepancy - whether it is in the manner of the statistical data accumulation or reporting, or whether it is in the application of the game rules themselves, including the "max" rules - especially in regards to baserunning.

The first attachment is a printout of the team Misc page after 12 games, for the Montreal Expos, in the ATG8 league "NL Round Robin Series #9". My handwritten tallies are on the sheet, as well as on the attached excel spreadsheet. Please note that my calculation of the total opportunities for Andre Dawson after 12 games is 7 baserunning opportunities. The 7th tick mark is a little faint on the PDF file. The details of his 7 baserunning opportunities are itemized on the spreadsheet.

Strat's stats show 20 baserunning opportunities, 9 advances, and 2 outs, after 12 games, per the PDF file.

My calculations are 23 baserunning opportunities, 10 advances, and 2 outs, after 12 games.

The discrepancy is 3 opportunities, and 1 advance.

The discrepancies relate to 2 baserunning opportunities for Andre Dawson, and 1 for the pitcher, Steve Rogers. Dawson has one more advance, per my calculations, than does strat-o-matic's stats.

I would like strat-o-matic to review this for me, and let me know what the discrepancies might relate to. The one for the pitcher seems straightforward, as strat's stats appear to exclude the pitcher baserunning.

The 2 discrepancies relating to Andre Dawson are difficult to reconcile. One theory might be that, in some cases, or on a random basis, Strat-o-matic might be leaving a SINGLE** as it is in the normal rules, instead of applying the "Max" rules - again, perhaps on a random basis. That might explain the discrepancy in "advances", since Dawson scored once on a SINGLE** result. It's impossible for me to know whether or not this is the explanation. I would greatly appreciate STRAT confirming this for me, if I'm right. If I'm not right, I would really appreciate the actual explanation for the discrepancy.

Apart from the discrepancy in the advances for Dawson, there seems to be an extra other opportunity discrepancy for Dawson, as I showed 7 opportunities for Dawson, whereas the strat Misc page shows only 5 opportunities for Dawson. Again, I would really appreciate it if strat could explain this discrepancy. The details of Dawson's baserunning opportunities are on the excel spreadsheet, but I'll copy and paste them on to this email as well:

Expos
Running opportunities

Games 1 through 12

Game // Opportunitiess // Advances // Outs // Runner advancing // Notes

1 // 0 // 0 // 0
2 // 4 // 1 // 0 // Ramirez // includes a SINGLE** converted to an opp (batter took second on throw), and a ballpark single (Pitcher, Rogers didn't advance)
3 // 3 // 2 // 0 // Cabrera, Dawson // includes a SINGLE* converted to an opp, and a SINGLE** assumed to be converted, and a SINGLE ** known to be converted
4 // 2 // 1 // 0 // Deshields // includes a SINGLE** converted to an opp (runner held)
5 // 0 // 0 // 0
6 // 1 // 1 // 0 // Dawson
7 // 0 // 0 // 0
8 // 0 // 0 // 0
9 // 8 // 4 // 1 // Ramirez (3) Dawson // includes 2 ballpark singles, a SINGLE** converted (runner was thrown out), and a SINGLE* converted to an opp.
10 // 3 // 0 // 1 // includes a ballpark single, and a SINGLE** converted to an opp (runner thrown out)
11 // 2 // 1 // 0 //Dawson
12 // 0 // 0 // 0

Totals 23 // 10 // 2

per strat Team MISC 20 // 9 // 2

Discrepancy 3 // 1 // 0

The discrepancies on the TEAM MISC stats are - Rogers (pitcher) one opportunity, no advance; and Dawson, 2 opportunities , 1 advance

Dawson

game 3 - Dawson scored from second on a SINGLE** hit by Carter. Carter held at first
game 4 - Dawson held up on third base (one base advance), on a SINGLE** hit by Deshields.
game 6 - Dawson advanced from first to third base, on a SINGLE by Carter - no asterisks
game 9 - Dawson held up at second base, on a ballpark single by Guerrero
game 9 - Dawson scored from second base, on a SINGLE* by Guerrero
Game 10 - Dawson thrown out trying to score from second, on a SINGLE** by Cabrera
Game 11 - Dawson scores from first on a double hit by Guererro

Total - 7 opportunities, 4 advances , 1 out.

Per MISC page for strat - 5 opportunities, 3 advances, 1 out

discrepancy - 2 opportunities, 1 advance

The URL for the individual strat game pages are as follows:

http://onlinegames.strat-o-matic.com/game/424625/4

http://onlinegames.strat-o-matic.com/game/424625/10

http://onlinegames.strat-o-matic.com/game/424625/16

http://onlinegames.strat-o-matic.com/game/424625/22

http://onlinegames.strat-o-matic.com/game/424625/28

http://onlinegames.strat-o-matic.com/game/424625/34

http://onlinegames.strat-o-matic.com/game/424625/39

http://onlinegames.strat-o-matic.com/game/424625/45

http://onlinegames.strat-o-matic.com/game/424625/51

http://onlinegames.strat-o-matic.com/game/424625/58

http://onlinegames.strat-o-matic.com/game/424625/64

http://onlinegames.strat-o-matic.com/game/424625/70

Thanks very much!

Hitters (14) min. 13, max. 17 (total min. 24, max. 28) Ballpark FX Clutch Rolls SB-CS Baserunning PH
Name B P GB-FB GIDP GWBI SI Opp HR Opp Hits Outs Hit Pit 2B 3B Hm Opp Adv Out AB H HR Salary
Carter, G. (1980) R C 22-18 3 0 0 1 1 4 0 0 26 20 1-0 0-0 0-0 1 0 0 0 0 0 6.58M
Pocoroba, B. (1982) S C 0-1 0 0 0 1 0 0 0 0 1 0 0-0 0-0 0-0 0 0 0 1 0 0 .91M
Hassey, R. (1990) L C 0-0 0 0 0 0 0 0 0 0 0 0 0-0 0-0 0-0 0 0 0 0 0 0 .69M
Galarraga, A. (1988) R 1B 16-15 0 1 0 1 2 4 0 0 22 28 1-0 0-0 0-0 1 0 0 0 0 0 6.07M
Deshields, D. (1997) L 2B 20-17 0 0 0 2 0 0 0 0 24 19 2-2 0-0 0-0 1 1 0 0 0 0 1.96M
Scott, R. (1980) S 2B 3-0 0 0 0 0 0 0 0 0 2 2 0-0 0-0 0-0 0 0 0 1 0 0 .85M
Mota, M. (1962) R 2B 0-0 0 0 0 0 0 0 0 0 0 0 0-0 0-0 0-0 0 0 0 0 0 0 .50M
Cabrera, M. (2006) R 3B 15-24 1 2 1 2 1 4 0 0 27 25 3-0 0-0 0-0 5 1 1 0 0 0 6.97M
Ramirez, H. (2008) R SS 22-18 1 1 2 3 0 2 0 0 28 26 1-0 0-1 0-0 6 4 0 0 0 0 5.63M
Raines, T. (1986) S LF 20-12 1 0 1 3 0 0 0 0 26 14 4-0 0-0 0-0 0 0 0 0 0 0 7.29M
Leflore, R. (1980) R LF 5-5 0 0 0 2 0 0 0 0 8 11 2-1 1-0 0-0 0 0 0 3 0 0 .77M
Dawson, A. (1980) R CF 23-11 0 0 1 2 0 0 0 0 26 19 0-1 0-0 0-0 5 3 1 0 0 0 7.21M
Guerrero, V. (2002) R RF 20-17 0 4 1 1 1 3 0 0 25 28 3-1 1-1 0-0 0 0 0 0 0 0 9.53M
Cromartie, W. (1982) L RF 2-1 0 0 0 0 0 0 0 0 2 3 0-0 0-0 0-0 1 0 0 4 1 0 1.18M
TOTALS 183-143 6 8 6 18 5 17 0 0 231 207 17-5 2-2 0-0 20 9 2 9 1 0

As difficult as it is to read, this is the copy and paste from the team Misc page after 12 games, just to show the historical reference for the strat statistics.

Here's the link to the actual page (although it will keep getting updated, of course):

http://onlinegames.strat-o-matic.com/team/misc/1384760