Minor to the Majors: More on Pitcher Grades & Control

A couple of weeks ago I made an initial stab at projecting a pitching prospect’s MLB value knowing their prospect grades. I failed miserably, but got some ideas in which to move forward on. Today, I will test some of those theories and see if I can tease out any information.

From the previous study, I found the following on pitcher grades.

1. Pitchers who are a few years from the majors are likely to change quite a bit before for they reach the majors. They could have arm surgery, find command, or add a pitch. Early grades have too much noise to be used for future results.

2. Fastballs are almost exclusively graded only on velocity. For a soft-tossing prospect to get a major call up, their fastball is likely better than just the velocity shows. They likely have plus movement and/or command of the pitch.

3. Almost all pitching prospects will get stuck with a future 50 control grade because getting graded less than 50 is a death sentence for their progress. Not all major league pitchers have average control. Exactly half don’t, so the control grade needs to be adjusted.

4. From some discussions with people in the game, I found that having one or two plus breaking balls can help carry a pitcher with an average fastball.

With these limitations, I decided to relook at pitchers by implementing the following guidelines for this study.

1. Look at pitcher grades who were in the high minors and called up the year they were graded.

Implementing step 1 is fairly simple and I took the data a step further. When looking at major league results, I only looked at results two years into the future. Pitchers change so much that looking ahead is futile. Once a pitcher has been in the majors a couple of seasons, fantasy owners should have a decent idea of the pitcher’s talent and prospect grades will no longer be needed.

2. Ignore all fastball grades

Nothing needs to be done with ignoring fastballs, but to ignore them. Done.

3. Use the pitcher’s high minors walk rate

With the walk rates, I used the pitchers’ walk rates from Double-A and Triple-A. I figured I would need to make an adjustment to the Double-A value to project it to Triple-A where their results should worsen. What I found was that the walk rate was basically flat between the two levels with some pitcher improving and other getting worse. More on this later.

Additionally, I tracked their walk rate in their first major league season for a comparison.

4. Track the number of plus breaking balls or changeups (graded 60 or higher).

This was simple to follow with pitchers getting a single integer value.

Determining Control

Using the above factors, I collected the grades from Baseball America from 2011 to 2015. While Baseball America graded quite a few pitchers, only 40 pitchers met the above criteria. This sample is looking more at the quality of the data versus quantity.

I am going to start with the control values because I was able to find some encouraging results. Just as a reminder a 50 grade is supposed to mean league average control and a 60 grade is one standard deviation better. Here are the grades with the actual MLB benchmarks for each grade.

Taking the pitcher’s Control grade and minor league walk rates, I compared them to their major league values. Just looking at the Grade versus actual walk rate, there was a definite relationship, but the r-squared value was only 0.16. Additionally the prospect grade (from Baseball America) was on average 7 points higher than the actual results using the above table. So if a pitcher had a 50 grade (3.1 BB/9) on Control, their results were closer to a 43 Grade (4.0 BB/9).

Now just using the average walk rate from Double-A and Triple-A and comparing it to the major league walk rate a better correlation exists with an r-squared of 0.22. Additionally, these prospects had their high minors BB/9 jump by .22 when they got to the majors.

Bringing the two values, the grade and minor league results, together produce the best final value using a few different methods. I ran into a little bit of an issue because I wanted a simple solution to remember and none of the answers fits this mold. Here are three possible methods.

For all three, convert the pitcher’s Control Grade into a walk rate using the above table.

Next find the pitcher’s average walk rate from Double and Triple-A.

Use the above two numbers in one of the following equations.

• Average the above two values and then add just a bit to the high side of the walk rate.
• Take the average of the two values and put them in the following formula: 1.05 * (Averaged Value) + 0.2
• Use the following formula which weights minor league performance more: 0.60 * (Minor league walk rate) + .50 * (Grade walk rate)

As you can see, the second and third formulas are really close to the first one, but just a bit lower.

For an example, here is how the numbers would work out for Lucas Giolito.

2016 Control Grade: 55 or 2.7 BB/9
Triple-A walk rate: 2.4 BB/9
Double-A walk rate: 3.9 BB/9
Average: 3.15 BB/9

Results
Method 1: 2.9 BB/9 (no adjustment)
Method 2: 3.3 BB/9
Method 3: 3.2 BB/9

Those are the various values which all hover around 3.0. Giolito decided to completely lose control in his 21 innings pitched and he posted a 5.1 BB/9. Knowing his minor league track record, I would not be surprised to see his walk rate drop a bit if he pitches next year in the majors.

Well, that is it for this week. I have been able to go through the new setup and report some decent results for the Control aspect of a pitcher’s prospect grade. Next week, I will report the results of above average non-fastballs and take that final step in trying to project pitching prospects. In the meantime, let me know if you have any questions.