Using OXS to predict runs

on 05/20/2002


There has been a lot of debate about OXS, is ability to determine a players worth, and its ability to predict "runs scored." Much of the discussion stems from confusion about the formula itself and whether or not it is "98%" accurate. Because people don't understand how it works and its results, they don't understand how it can be used or if it is even useful at all.

Still more confusion (or even distain) results from AB x OBP x SLG's biggest proponents. By expounding its virtues at every opportunity, people sometimes get the mistaken impression it is the only personnel evaluation tool worth using. While not true, this mistaken impression is natural enough.

Hopefully this article will make clear how OXS predicts runs and how that predicted total can and should be used.

First of all, what exactly is this formula?

AB x OBP x SLG = Predicted Runs Scored

Pretty simple, huh? That is one of the best features of using OXS to predict runs.

While one can use OXS with individual players to calculate predicted runs, it is best used with TEAM totals and averages. In other words:

Team ABs x Team OBP x Team SLG = Total Runs Scored by the Team

The fact that OXS is most accurate when the numbers are team orientated seems to confuse a lot of people. Now, that doesn't mean that it can't be used when evaluating players. More on that later.

What is the "98%" accuracy claim?

This probably has caused more "discussion" than any single issue on Brewerfan.com, with the possible exception of Ron Belliard.

I have no idea how that number was derived, and I won't say that the formula predicts runs within 98% each and every time. Simply stated, when taking a teams ABs x OBP x SLG, the result will "predict" how many runs that team scored during that season within about +/- 2% (hence the accepted figure of 98%). Here are some examples, using our own beloved Milwaukee Brewers. (all stats courtesy MLB.com)

YearABRunsOBPSLGPredicted RunsPercentage

As you can see, ABs x OBP x SLG is not perfect, but one shouldn't get hung up on the fact that sometimes it is only 96.7% accurate. One is able to consistently predict within about 3% how many runs a team will score using this method.

Better still, the results seem to be nearly timeless. One can use the same formula for the 1927 Yankees or the 1930 Cubs or the 1941 Indians and figure the results in a few seconds.

Why use this formula?

There are definitely other formula's available that take into account other facets of the game, such as the stolen base. I probably don't qualify as a "stats geek" because I don't know anything about those other tools. That is why ABs x OBP x SLG that is popular: it is easy to use. It is this ease of use that makes it a useful tool, particularily for those without a degree in calculus.

So, before we get into how OXS can be used, lets review the basics.

AB x OBP x SLG is best used to calculate how many runs a team will score during a given period, most often over the course of a season, based upon three simple components: How many ABs a team had, their team OBP and team SLG. While AB x OBP x SLG can be useful to give value to a single player, it is better used to predict team runs over the course of a year.

Real Life Baseball Applications of AB x OBP x SLG = Predicted Runs

Scenario 1:

During the annual postseason evaluation of his team, a GM and his staff are trying to determine why his team didn't score more runs than they did (723). Everyone feels like they should have done better, but for some reason the offense never got untracked.

Using the formula AB x OBP x SLG he calculates his team should have score 715 runs. This means his team scored 1.1% more runs than they "should" have using this formula.

How is this useful? What can the GM and his staff do with this number?

Evaluation 1a:

Well, taken in isolation, he can relatively quickly determine whether or not his team met realistic offensive expectations. "Wow, I guess our offense performed about as well as can be expected. Maybe we are not as good as we thought, guys? We need to get some more productive players next year or we can expect the same kind of thing." However, no GM worth anything would stop at this point.

Evaluation 1b:

What if the team had a major injury? Well, that same GM could say "Well, you know, we normally get a lot more production from our SS, but seeing as how Jason Caroway tore his ACL in spring training we didn't get that production. I bet that was the problem." The GM could then substitute the production he got from his SS with typical numbers for Jason Caroway, recalculate his teams OBP and SLG, and re-run ABs x OBP x SLG. If the team should have scored 730 runs with Caroway playing short, he knows that a return to health of his everyday SS should help the team. He could also see that Jason Caroway would not solve the offensive woes entirely.

Evaluation 1c:

Seeing that his team met expectations, and that the return of his All-star SS would only improve things to a certain point, the GM could substitute the numbers of potential FAs (or players he'd like to traded for) and see how their production might have an effect on the offense. Seeing as how he only needs to substitute a few numbers, he can pretty quickly see if he is on the right track. After all, you don't want to sign a player to a big contract and find out no matter how well he plays your team might still not be very good on offense.

Notice that in this example the GM is using ABs x OBP x SLG as a TOOL to HELP evaluate his current and potential future rosters. He still needs to scout FAs before signing them, determine how much he has left in his bat, whether his team can afford him, etc. He still needs to factor in things like age and look at whether or not guys had career-years. The addition of ABs x OBP x SLG to the GMs baseball knowledge could help prevent personnel mistakes that might eventually lead to his dismissal later.

Scenario 2:

Much like the GM in Scenario 1, the GM in this scenario is doing his annual post-season evaluation. The team's unexpected rise to a wild card berth caught everyone - fan, media and GM alike - by surprise. Fueled by an offense that was much more explosive than anticipated, his club earned the leagues wildcard berth and was defeated in a heart-wrenching five game series.

During the evaluation, someone on the GMs staff runs the numbers for AB x OBP x SLG. The GM and staff are surprised to see it calculates 741 runs, instead of the 803 scored, meaning the produced about 8% more runs than expected.

Evaluation 2a:

The GM and the staff discuss the numbers. How many games did they score a number of "odd" runs last year? The media said all year that this was a "team of destiny". Maybe this team will still be good- after all there is a lot of young talent out there. But, at the same time, maybe they were a little bit lucky last year on offense. They can't count on being lucky next year, so steps can be taken to help improve the club.

Now, the GM still has to figure out what players to keep and which ones to replace. While he can use statistics to help him, he still has to evaluate his players like he has in the past. Using OPS (OBP X SLG) and OXS to help evaluate those players and his team, will allow him to make smarter personnel decisions.

For the fan, OXS and OPS are a means of quickly and effectively evaluating offensive players without working for a MLB organization. There are not many of us who are lucky enough to see the hundreds of games necessary to make solid judgments about every player that comes up in baseball conversations. Again, there are other ways of determining a player' s value. And, so far as I am concerned, no statistic will ever make up for seeing a player in person repeatedly. In lieu of such, I think OPS and OXS are the best tools around.


As you can see, AB x OBP x SLG is just a tool, not unlike a cordless screwdriver. You can't use it exclusively when making personnel decisions; just like a cordless screwdriver isn't the only tool you'd use to build a house. Like a cordless screwdriver, AB x OBP x SLG is a versatile and powerful tool that a lot of people wouldn't want to be without.