I Created My Own QB Metric, Tell Me What You Think

[fieryprophet]

Called QBA (Quarterback Assessment) or "Cuba"

http://50.116.44.142/qba/

For you nerd types like me, this is the formula:

plays = attempts + rushes + sacks

positive_plays = completions + rushes

total_touchdowns = passing_touchdowns + rushing_touchdowns

total_yards = (passing_yards + rushing_yards) sack_yards_lost

turnover_yards = (interceptions * 50) + (fumbles * 25)

adjusted_yards = total_yards - turnover_yards

expected_touchdowns = floor((total_yards + turnover_yards)/75)

adjusted_yards_per_play = (adjusted_yards/plays) // also known as APP

total_points = (total_touchdowns * 7)

Efficiency % = ((positive_plays/plays) * 100) //cannot be less than 0 or more than 100

APP % = ((adjusted_yards_per_play/10) * 100) //cannot be less than 0 or more than 100

Offense % = ((adjusted_yards/team_total_yards) * 100) //cannot be less than 0 or more than 100

TD % = ((total_touchdowns/expected_touchdowns)*100) //cannot be less than 0 or more than 100

Score % = ((total_points/team_score)*100) //cannot be less than 0 or more than 100

Opponent % ((total_points/opponent_score)*100) //cannot be less than 0 or more than 100

QBA = (Efficiency % * 0.225) + (APP % * 0.225) + (Offense % * 0.225) +

(TD % * 0.225) + (Score % * 0.1) + (Opponent % * 0.1) //cannot be less than 0 or more than 100

Don't mind the outliers like Matt More, he played one game and it scored very well. (in b4 TRD)

[SgtJoo]

I'll tell you what I ESPN: Stop trying to make QBR happen.

[fieryprophet]

I'll tell you what I ESPN: Stop trying to make QBR happen.

At least look at it. You can't see the numbers that go into QBR, you can with this.

[Thorrez]

I admire you effort.

What I would like to see is more stuff that mater for the outcome of the game. A pass on 3rd and 8 is (almost) as bad with a completion for 7 as an incompletion. 8 yards and a first is great. Yet, 7 and 8 yards will show up the same in this measurement.

Kind of like our inflated total rushing yards under Fox (with a lot of useless 7-8 yards rushes on draw plays on 3rd and 10+).

[fieryprophet]

I admire you effort.

What I would like to see is more stuff that mater for the outcome of the game. A pass on 3rd and 8 is (almost) as bad with a completion for 7 as an incompletion. 8 yards and a first is great. Yet, 7 and 8 yards will show up the same in this measurement.

Kind of like our inflated total rushing yards under Fox (with a lot of useless 7-8 yards rushes on draw plays on 3rd and 10+).

That would require poring over a drive chart and defeat the design goal of the stat to be able to be generated by a box score. It also would require interpreting intent on those plays in question. Sometimes a dump off is absolutely the right play.

[TbTeRRoR]

Yes and to if you dump it off and a back breaks two tackles and gains 20yds how is that a triumph by the QB? The dump off is supposed to be a gimmie then the back does all the work/ with blocking help of course.

[fieryprophet]

Yes and to if you dump it off and a back breaks two tackles and gains 20yds how is that a triumph by the QB? The dump off is supposed to be a gimmie then the back does all the work/ with blocking help of course.

You just described a screen play. It still requires the QB to execute it by fooling the defense long enough to get the RB in position.

[TbTeRRoR]

You just described a screen play. It still requires the QB to execute it by fooling the defense long enough to get the RB in position.

Yes I know. I was responding to the 3 and 8 and the QB passes for 7. You would want metrics to show that as something that lowers his rating and yet a pass thrown under the sticks that goes for long yards would help the QB? Is that good? The point being you can only measure so much a QB does. The skill set of those he's working with can and will factor in. Case in point: Cam did a good job on the dump to Deangelo (although it was tipped) But that play could have went for 5 yds or a TD. Which is completey out of the QBs hands.

[Varking]

Its not bad!

[Marguide]

It passes the eyeball test pretty well. The guys that have played well rate highly, and that is the most telling part of any QB rating.

It certainly looks better than what ESPN is promoting.

[TANTRIC-NINJA]

I would definitely add a swag metric...Cam = 438, RG3= 120, Luck = .038, Alex Smith = - 230, Mark Sanchez = - 17 million as in Cap Hit!!

[mav1234]

ESPN is trying to promote something that actually takes into account what the situation in the game is like, etc, but I think that can be kind of subjective.

It's interesting... not sure how much more useful it is than any other QB rating, but it is cool that we can easily see what is going into it and it is more useful than a purely passing metric in today's game. Props for this!

[philp]

Pretty cool man. I wish my phone liked your site though. I'll check it out when I hop on my computer.

[fieryprophet]

It passes the eyeball test pretty well. The guys that have played well rate highly, and that is the most telling part of any QB rating.

It certainly looks better than what ESPN is promoting.

I thought so as well. For one thing, unlike passer rating it analyzes a QB's performance based on his whole team's performance. If they have a monster rushing day and tons of points scored it will be weighted against a QB who didn't create a lot of those yards or points.