Colorado Rockies Win Predictions — SABR Style


Last night watching the Rockies play the Indians was a real treat. 20+ hits, 18 runs — smash mouth baseball! Who doesn’t love that? Brings me back to my days of Little League, watching the Blake Street Bombers. It gets a little too surreal when I see Dante and Walt on the bench. Jeff Francis looked pretty solid last night, as he has all spring, even if at times he was a little too hittable. That is just what you get with him. When he locates he is pure joy, when he leaves it over the plate he is pure anguish.

Lately I have been harping on the Rockies pitching and who can really blame me? The pitching is the crux of this season to the point that the flux in wins between a 4.30 team ERA and a 4.70 team ERA is up to plus-minus 10 wins depending on how many runs the Rockies score. (Pause while I high-five myself for using crux and flux in the same sentence.) That is a huge difference! In my opinion that could be the difference between being a contender for a Wild Card spot or being a seller in July. This team has no questions about offensive firepower, especially if it can stay healthy. The question for the 2013 Rockies is if the pitching can keep the offense relevant! As we come to the end of Spring Training I wanted to take the chance to give you my win prediction for the 2013 Colorado Rockies.

I know some of you are saying, “WHOA where did this guy come up with +/- 10 wins?”. I am so glad you asked and I would love to explain, so let’s dive in.

Bill James is both loved and hated in the baseball community because of his eccentric outlook on baseball and it’s ability to be predicted. He is considered by most to be the Grandfather of Sabermetrics, or as most of the popular world knows it — Moneyball. These metrics are not near as complicated as one might believe, so try not to go into Oh-My-God-He-Just-Said-Math mode. Mostly because I haven’t even said math…yet. Ok let’s bring in the math. One of James’ many equations was a basic Pythagorean Expectation to calculate a teams success. Wins can be calculated based simply on the amounts of runs a team would score versus how many they give up. So simple. In true John Madden-esque logic the team that scores more runs than it allows wins more games! It is not always the case, sure, but it does serve as an excellent proxy.

This year the Rockies have been projected by most sports books to end the 2013 season at 71.5 wins, last in the NL West. In light of this RoxPile editor Hayden Kane wrote a wonderful article a while ago pushing all our betting faithful to take the over, and I agree whole-heartedly. Here is the logic:

Baseballs. Image: Douglas Jones-USA TODAY Sports

The Rockies in the last three seasons have plated an average of 754 runs per season according to FanGraphs. At this point I want to introduce another metric to the Rox Pile faithful called weighted Runs Created (wRC). This metric is simply showing how many runs the Rockies should have scored in a season, or in essence what bad luck they had. The Rockies’ wRC over the past three years was 765 runs/year. Not much different than actual so not a ton of bad luck (or underachieving). I am going to use that number as my baseline to test this theory even though I believe the actual runs scored in 2013 will be much higher as the past few seasons have seen the Rockies without both Cargo and Tulo for significant periods of time. But that’s fine, we’ll use 765 runs.

Now we bring in the pitching aspect. Over the last three seasons the Rockies pitching has thrown up a 4.59 ERA. Not horrid especially considering the monstrous 2012 ERA of 5.22. A 4.59 ERA equates to 744 runs against per season. Using Bill James’ win prediction formula to establish the baseline, Wins = 1/(1+(RA/RS)^2)*Games = 1/(1+(744/765)^2)*162 = 83 wins/season. In actuality the Rockies averaged 73 wins per season. That’s a pretty big gap from reality, yeah? I agree. Lets adjust for this difference by applying a factor of 0.90 to whatever win prediction we come up with. That is a good safety factor and accounts for the general inaccuracy of using runs to determine wins.

In 2013 I believe the Rockies will have an offensive outburst, to the tune of 4.9 runs/game. That equates to 794 runs (In 2009 they had 804 runs, for reference) scored over the season. I also believe that the Rockies pitching, though still not good enough, will be better than last year. I certainly feel it could have been much better if the Rockies had made a couple of moves not named Jon Garland, but using the average ERA of 4.60 feels right. This gives us a predicted 77 wins for 2013.

There is also a weighted FIP value (everyone knows how much I love FIP right?) that shows how many runs the Rockies should have allowed. This xFIP metric for the last 3 seasons is 4.02 runs/game.. that would be a bad-luck free prediction of 87 wins this season. Now immediately disregard this value because anyone who reads this knows no baseball team is free from bad luck.

I like the 77-win prediction, but it is not going to be enough to get us into the playoffs. Wait — is it even reasonable to think we could be in the playoffs come October? Why not? What would it take, pitching-wise, to get us there? In the past few years it has taken 90 wins to get to the post-season so leaving the offensive output the same (794 runs), the Rockies would have to allow only 625 runs to get to that magic 90 win level. This equates to a 3.85 team ERA… ok so they probably won’t make the playoffs this year. But I truly believe 80 wins is an achievable mark in 2013, and I would feel pretty good if they could play to that level. I mean, .500 baseball out of this team? It’s right within their grasp.

Last night my wife asked me if it I was allowed to write that many negative outlooks on the Rockies. I replied that it absolutely was because I’M A MAN! I’M 25! But at the end of the day I am as much an optimistic fan of the Colorado Rockies as you all are. I just use math to back me up.