Run differential

I've always thought it was a irrelevant stat, yet it's used as the tiebreaker for many SOM related leagues / tournaments. Case in point, just finished Tour Event 5 league, my run diff was -10, yet I had the best record in the league.

http://onlinegames.strat-o-matic.com/league/427695

JT

We had this discussion in a league awhile ago and MarlinGene (who is a self-admitted math nerd) pointed out:

Even though many professional sports analysts use the term "differential", it's incorrect.

The key point is how it is calculated. If it is subtraction, the result is called a "difference". A "differential" is a component part of a Calculus derivative.

From then on, I promised to point it out whenever I saw it...

I wouldn't call it an irrelevant stat--it generally indicates the strength of your team. Now, if you are great at winning 1 run games and/or your team is blown out more than you blow out the other team then your team could perform better than it's run differential would suggest. Your team was 24-17 in one-run games and it was 15-22 in games decided by 5 runs or more. So those figures might explain a good part of why you out-performed the run differential.
Of course, all that matters is that you win, not the run differential, but I guess I would prefer to have a +100 differential team than a negative one. I am guessing that a 90 win total is about as high as it gets for a negative run differential team.

From the tour through event 4, 8 of the champions had a greater than +100 differential, and another 14 were greater than. 50. There were 4 champions with negative differentials, but their win totals were 84,83, 85 and 91--there was only champion with a win total way out of line with its run differential.

So run differential is not determinative...it's not irrelevant either.

The sabermetric proponents of expected winning percentage would argue your best record was a function of good luck.

Plus we have the question in those specific situations where run differential is used what tiebreaker method would you propose?

Run differential can be mis-leading, as if your blowing out teams it will skew the value. As a tie-breaker, your gaining n advantage by scoring as many runs as possible. So not all wins are equal.

I would like to see instead of run differential, a statistic of median run differential being used, so the blowouts do not impact this. Downfall I see with this would be 1-run games being dealt a blow as well. So perhaps median run differential of say the lowest 80% or something like the maximum run differential be like 5, so blowouts count just weighted differently.

Instead of tie breaker have a 1 game playoff like MLB would do.

If the object of the game is for your offense to score runs and for your pitching to not allow runs, then run diff is about the simplest short-hand method I can think of to reflect any teams' expected performance. It correlates with winning pretty well. If you knew nothing else about a roster except for their final run diff, you would choose the roster with the higher one every single time.

Now, in SOM Online, you can have leagues with ten pitcher's parks or leagues with an equal talent level of managers and there can be all sorts of reasons for teams to win with low run diffs. But you would never SEEK OUT to be a "low run diff team." It's completely counter to the object of the game (i.e., scoring more and allowing less).

Regarding the tour tiebreaker...

The tiebreakers for the tour are rings, finals, playoffs. If run diff is the fourth tiebreaker, I don't think that's anything crazy.

While I will agree that one would seek out team with best run differential, I have to disagree with the object of the game being for you offense to score runs and for your pitching to not allow runs, in regards to blowouts as if a team is getting blown out, who is pitching? Your worse RP most of the time and will stay in game as long as Hal will allow, but your offense does not change, you do not see subs in SOM online like you do in MLB, so run differential has its flaws.

The goal is to win the game. How often does your top RP or closer enter the game in a blow out, if there are other arms available? Once the same is a blow out, then the object of the game changes, it goes to not wasting your bullpen and hoping no one gets injured.

Wow, all good points, arguing both sides. It's refreshing on the the board now-a-days to have a quasi-intellectual discussion on a SOM topic. Thank you gentleman.

JT