# Honest evaluation

Two days into the regular season, hopefully you still are optimistic of your team’s chances this season. Most people are, actually, and the number that are is usually too many. Sometimes all people need is a slight reality check to make them swallow their egos and put them in their proper place. I am not suggesting you are one of those guys, but why not do a fairly simple yet somewhat time consuming activity to make sure?

The basic gist of the activity: Get all of the projections for the players on your fantasy squad from a projection system (THT’s projections, CHONE, Marcel—your choice) in a spreadsheet, calculate the projected totals for your team in each stat category, and then compare your team’s results to the other team’s results in your league.

Obviously I do not believe that whatever projected totals this process spits out are necessarily the way the league standings will look at the end of the year; injures, midseason roster adjustments, and projection systems not being 100 percent accurate are responsible for that. But this exercise should hopefully give you a fairly accurate, unbiased evaluation of your team compared to the others in your league, maybe changing your perspective on your team.

### Process

In case you are not sure exactly how to go about this, I’ll walk you through the process, step by step.

First, we have to pick a projection system to use. As I mentioned before, you have many choices here. You can use THT’s projections (available for $10) or any of the projection systems FanGraphs features at their projection page.

Once you’ve got the projections in a spreadsheet, you have to manually group the players and their projections in separate sheets by team. This can take some time, so if anyone knows a better way, feel free to let us know.

Getting the team totals for stats like home runs and strikeouts should be simple enough; all that is required is simple addition. For rate stats like batting average and ERA, however, things get a bit more tricky. Simply averaging the player’s projected batting averages together is not a good idea because a player with a .300 average over 600 at-bats should have a greater effect on the team’s batting average than a player with the same average over 250 at-bats. Averaging the batting averages would not account for that.

What you must do instead is divide the total team hits by the total team at-bats. If you are using CHONE projections, this is not too much of a problem as hits and at-bats are given for each player. What if hits are not given? Simple. Do batting average multiplied by at-bats. What if at-bats are not given (as they are not in the THT projections)? Simple. Do hits divided by batting average.

The same process works for ERA. Instead of averaging ERAs together, take the total projected earned runs allowed divided by the total projected innings pitched for your team. To get ERA from that you have to multiply by nine, by the way. If earned runs are not given, you can get earned runs from ERA by multiplying ERA by innings pitched and dividing by nine, always.

I think most of you understand the process by now, if you did not already before, but I’ll spell out WHIP as well for those who need it. Similar to the others, it is total hits allowed plus total walks allowed over total projected innings pitched.

### Results

So now at this point you should have one projected number for each stat, for each team. Creating a new sheet with all of the team’s projected totals stacked up so they can be sorted and compared is a smart idea. How does your team compare? Ideally the areas you are weak and strong in become evident.

It can be frustrating if your team is lower than you think it should be in, let’s say, ERA because you used CHONE projections and a team ranked just ahead of you owns Javier Vazquez and his projected 3.26 ERA and 200 strikeouts. You are thinking, “No way Vazquez has a 3.26 ERA! I mean I like the guy this year, but an ERA around 3.60 to 3.70 seems more reasonable to me…”

I have no problem with making slight adjustments like that, but do not go crazy. Projections systems do pretty well for themselves and most certainly are more accurate than whatever personal rankings you developed on your own. This does not mean I think you should value players strictly based on projection system projections; I do not because there is always a group of players I develop strong opinions for and those are the players I target or avoid.

One thing to keep an eye out for—and this would be really cool if pulled off—is teams that are projected to be unusually good. What does this suggest? That whoever drafted the team might have used the same projections system you are using now to draft their team. That knowledge can help you find out what players this person might overpay for in a trade and is also something to keep in mind for next year’s draft (are we already talking about those?).

### Application

In my last Roster Doctor article, there was some disagreement as to whether my favorable assessment of the team’s pitching staff was correct. One thing we can do is evaluate the team using the method I just described. Using the THT projections, I get the following projected totals for the team:

+-----+-----+-----+----+-------+----+-----+------+------+------+ | R | HR | RBI | SB | AVG | W | SV | K | ERA | WHIP | +-----+-----+-----+----+-------+----+-----+------+------+------+ | 834 | 222 | 845 | 96 | 0.292 | 71 | 120 | 1041 | 4.00 | 1.31 | +-----+-----+-----+----+-------+----+-----+------+------+------+

__Note:__ These pitching stats assume the team starts Vazquez, Baker, Meche, Parra, Price, and the closers.

__Note 2:__ These pitchers project for a combined 1096 innings, which makes the K/9 an impressive 8.54.

Those numbers alone mean very little. And without knowing the rosters of the other teams in the league it is going to be difficult to establish a context. Luckily, this league is close to a standard Yahoo league, and every year Yahoo comes out with a nice article in which they give the average stats of team that finished in the top three and the average production they received from each position.

Unfortunately the league of the team I analyzed has an extra UTIL position, so the comparison will not be exactly accurate. With ten starting hitters instead of the standard nine in most Yahoo leagues, hitters are more scarce, making the average production from each hitter comparatively lower in the Roster Doctor league. Understanding this flaw, here are the hitting results:

+----------------------+----+----+-----+----+--------+ | | R | HR | RBI | SB | AVG | +----------------------+----+----+-----+----+--------+ | Roster Doctor Team | 83 | 22 | 85 | 10 | 0.292 | +----------------------+----+----+-----+----+--------+ | Yahoo 1st Place Team | 94 | 25 | 92 | 18 | 0.294 | | Yahoo 2nd Place Team | 91 | 23 | 89 | 16 | 0.290 | | Yahoo 3rd Place Team | 89 | 22 | 87 | 15 | 0.287 | | Average Yahoo Team | 81 | 21 | 79 | 11 | 0.285 | +----------------------+----+----+-----+----+--------+

I am not sure how much the average production of a first, second, third, and average team in a ten-team hitting league differs from the numbers above, but given slight upwards adjustments to the Roster Doctor Team it is apparent its hitting is very good. I would expect mostly 3-2 and 4-1 victories from the hitting.

Now for the same chart only for the pitching:

+----------------------+----+----+-----+------+-------+ | | W | SV | K | ERA | WHIP | +----------------------+----+----+-----+------+-------+ | Roster Doctor Team | 8 | 13 | 116 | 4.00 | 1.31 | +----------------------+----+----+-----+------+-------+ | Yahoo 1st Place Team | 10 | 16 | 126 | 3.27 | 1.20 | | Yahoo 2nd Place Team | 9 | 13 | 120 | 3.43 | 1.23 | | Yahoo 3rd Place Team | 9 | 12 | 116 | 3.54 | 1.24 | +----------------------+----+----+-----+------+-------+

__Note:__ I could not determine what the average team was because while the Yahoo article gives the average production from a No.1 SP, No. 2 SP and so on, that does not mean that the average team will have one of each tier of SP. For hitters I could come up with an average because it is safe to assume every team will own at least one player at each position.

Looking at the above chart, it appears the commenters were correct; this team should be “on the wrong side of 4.00/1.30.”

Remembering that this is a K/9 instead of regular strikeout league, lets think of what would happen if this person were to only start his closers, Vazquez, and Baker. The wins category would be virtually punted, but now this team figures to be more competitive in both ERA and WHIP. Let’s re-run the numbers to see exactly what numbers are spit out:

+----+-----+-----+------+------+ | W | SV | K | ERA | WHIP | +----+-----+-----+------+------+ | 42 | 120 | 642 | 3.70 | 1.22 | +----+-----+-----+------+------+

The ERA is still somewhat high (thanks to Baker’s projected 4.41 ERA) but otherwise it looks very strong in saves, K/9, and WHIP. This group of six pitchers is projected to throw 632 innings, so the K/9 rises to an impressive 9.14. I believe there are 25 weeks in the MLB season, so that means these pitchers should total 25 innings per week, exactly this league’s minimum. In weeks where both Vazquez and Baker are only starting once, this team will probably have to start a third pitcher once to make sure they do not fall short of the minimum.

I am sure the other managers in this league are looking for the same kind of deal, but any trade that consolidates your starting pitching would be helpful. A trade involving Baker and Parra or Meche or Price for one better starting pitcher would be beneficial.

Starting only six pitchers—two starters and four relievers—projects to make this team above average in both the hitting and pitching categories and so I still like this team and feel it will have a good season.

Print This Post

For Average, ERA & WHIP, if you are in Excel, why couldn’t you use the “sumproduct” function on ERA and IP, or AVG and AB, and then either divide by IP or AB to get the weighted average for the category in question? Then you wouldn’t need to know the number of hits.

This is the first year I’m using projections, and when projecting my teams, I’m running into issues with saves.

They generally don’t seem to be projected very well (at least in Marcels & CHONE). Is there a trick anyone uses, or is it mostly an unfortunate byproduct of the projections in question?

For those that aren’t into the long version presented here, if you have your league history, you can use that to get an idea of where you really stand by doing the calculations for your own team and comparing them to the league history. If things aren’t looking good, you have a problem. If they are looking good, then you should be pleased.