View Full Version : AccuScore analysis: Chargers-Steelers

11-15-2008, 09:26 PM
AccuScore analysis: Chargers-Steelers
By Stephen Oh
Special to the Union-Tribune
Friday, November 14, 2008

AccuScore runs 10,000 computer simulations of NFL games, one play at a time, to project likely outcomes. Here is its assessment of Sunday's Chargers game against the Panthers.

It was not easy, but the Chargers got by the Kansas City Chiefs and are still in good position to win the AFC West. They have a tough match-up Sunday in Pittsburgh, but considering the Steelers are 6-3 and the Chargers are 4-5, AccuScore gives the Chargers a strong 44 percent chance of upsetting Pittsburgh. In 10,000 simulations, Pittsburgh is winning by an average score of only 23-21.


While virtually no one thinks Byron Leftwich is a better quarterback than Ben Roethlisberger, there are plenty of reasons Leftwich may be a better option this week. The combination of injuries and poor judgment has resulted in Roethlisberger having three straight horrible games. Leftwich played great in the second half to lead Pittsburgh over Washington in Week 9.

AccuScore gives Ben Roethlisberger a 75 percent chance of starting Sunday. The Steelers' chances of winning are actually slightly lower with Roethlisberger starting.

Steelers QBs.......Pit. win%...Cmp-Att....Pct....TD....Int...Sack



Despite completing a higher percentage of his passes and being more effective scoring touchdowns in the red zone, Roethlisberger's problems with interceptions and sacks (nearly twice as many as Leftwich in both categories) result in the Steelers being slightly better off with Leftwich.


It's hard to imagine, but LaDainian Tomlinson has been held under 3 yards per carry in four games this season. Only twice has he averaged more than 5. The Steelers have the second-best run defense in the league and are holding opposing rushers below 3.5 yards per carry.

After 10,000 simulations, LT averaged 52 yards on 19 carries, only 2.7 yards per carry. LT does not need to have a huge game, but in simulations in which he had at least 60 yards he averages over 4 yards per carry with a 75 percent chance of scoring at least 1 rushing touchdown. More importantly, San Diego wins 62 percent of these simulations by an average of three points.

However, in simulations in which LT is ineffective and held under 60 yards he is averaging a paltry 1.9 ypc and Pittsburgh has a solid four-point advantage.

Tomlinson rushing.............%Chance...YPC...TD...SD win%..Score

Baseline (19 car., 52 yds.)........– ..........2.7.....0.4.....44%........PIT 23-21

LT 60-plus yards......................33%.......4.2.....0.8.. ...62%........SD 24-21

LT under 60 yards....................67%......1.9.....0.2..... 33%........PIT 23-19

(Well you know what software engineers say about computers; GIGO - Garbage in, Garbage out. If computers were always right then the filthy Cheatriots would have been 19-0 last year. Still I have never seen an analysis like this before so it is kind of interesting. - mesa)

11-15-2008, 09:52 PM
Pretty interesting read. But around a 50% chance of winning...... I don't need a computer or 10,000 of them to tell me those odds. No to take anything away from your post.

Again I found it interesting.

11-15-2008, 09:55 PM
AccuScore runs 10,000 computer simulations of NFL games, one play at a time, to project likely outcomes. Here is its assessment of Sunday's Chargers game against the Panthers.

Did they run the simulations on the wrong matchup? :noidea:

11-15-2008, 10:03 PM
Did they run the simulations on the wrong matchup? :noidea:

I didn't notice that when I first read. Oh well it's what I would expect from a San Diego newspaper which apparently can't tell the difference between the Panthers and the Steelers. Hey maybe the Chargers will prepare for the Panthers instead of the Steelers.

11-15-2008, 10:04 PM
I know they run the Accu-score thing for every game like this. I don't remember what the overall record was for this last year...but it was fairly decent (at least when compared to the "experts picks" at your standard sports media like ESPN).

I know in the office pool, I picked the Steelers to win. I'm figuring a 7-10 point victory.

11-15-2008, 10:07 PM
How does this AccuPoint thing work?

11-15-2008, 10:15 PM
How does this AccuPoint thing work?

AccuScore forecasts the future of sports, just like the weather forecast predicts the weather. We use a highly complex simulation engine run off of a bank of super computers and then we analyze the results. The simulation engine was originally developed to simulate human evolution and no, I'm not making that up. The engine would simulate from scratch, 10,000 times. By analyzing the data, the probability of any event could be calculated such as genetic mutation, tribal migration, you name it. Complex computer simulation is used everywhere from predicting the weather, the stock market and it is even used to predict government coups. Now AccuScore has brought that level of technology and analysis to the world of sports....

....So how does it work? Well, for simplicity sake, AccuScore creates virtual players using over 30 different attributes then we put those players on a team which also has attributes such as coaching tendencies, then we let them play a game, one play at a time. We record every possible statistic then repeat the process, 10,000 times. The AccuScore simulation engine even factors in things like wind, rain, snow, field type, injuries. You name it, it's taken into account.


11-15-2008, 10:48 PM

It's kinda like Madden. :toofunny: