So gang, I ran a spreadsheet on the winning percentage of every starting QB in the league as of game five (I used Jake Locker instead of Matt Hasselbeck in this analysis), to see if there was any statistical advantage to starting a rookie (Russell Wilson) over starting a “veteran” that had backed-up another player for a significant amount of time (Matt Flynn).
The base line for this investigation is the overall career Winning Percentage (WP) for every starting QB, which breaks down numerically as slightly less than 60% (58.35% to be exact), so the numbers will be plus or minus that base. The basis works out to slightly more than a 9-7 (9.34-6.66) record. I broke it down into four categories for each example; career WP, since acquired WP, first year WP, second year WP, and the sample size for a perspective.
So if you’re ready here we go….
Let’s start with the All QB category and look at the numbers that most show some insight on our QB controversy; which QB gives us the best chance to win Wilson or Flynn.
If you are a QB drafted by the same team you play for you have a 2.21% to be better than average, it’s an 11 person sample size, and is enough to bump you up to a 10-6 record…barely (9.69-6.31) one point for Russ!
If you are a QB who backed-up another QB for one year or more you jump to a 3.46% advantage still a 10-6 record but closer (9.89-6.11…18 sample) One point for Matt!
First year starters as a whole did poorly in general, but the worst was QBs that backed-up less than a year -16.53% or a 7-9 record in a five player sample. Minus 1 for Russ.
More than a year as a back-up was a -2.72% (9-7 or more exactly 8.9-7.1) but if they were acquired from another team and a first year starter they were -11.17% or a 8-8 record(7.55-8.45)so minus 1 for Matt.
Second year starters did better than first but only QBs that never or backed-up less than a year were positive at 7.69% or an impressive 11-5 (10.57-5.43…only a 5 sample size) record. Plus two for Russ …….next year.
Now just for a frame of reference I did all first round draft picks starting as of game five , they don’t involve our situation, but give insight on the next section that does go to the root of this investigation.
All First Round Draft Picks (FRDPs) Career WP were just a little negative (-.4o%) so very close to overall career WP, but it’s a 23 out of 33 sample size so that does make sense.
Most interesting was first year starters did as poorly as they did considering they were some of the best QBs coming out of college.
The best was -2.25% (still a 9-7 record) and it was a 4 sample size of All FRDPs Traded, which included the likes of Rivers, Palmer, E Manning, and Cutler, so take it for what it’s worth, unusual trades of some very good QBs.
All FRDPs, first year as a starter had a -12.66% (a 23 sample) or a 7-9 record and all FRDPs that never backed-up a QB did even worst -15.6% (a 12 sample) yet still a 7-9 record.
The worst swing was All FRDPs Free Agents a sample size of two, Vick and P Manning’s first (-21.25% a 6-10 record) and second (21.65% and a 13-3 record) year starts. They also did quite well in Career WP *duh* (6% and a 10-6 record)FYI.
The only other interesting stat (IMO) was All FRDPs who never backed-up another QB, for their career had a 1.13% increase over the average or a 10-6 record (a sample size of 12).
Okay gang if you stayed with me for this long, here’s the meat of my statistical study…
All QBs Drafted 2nd or Below… Since both Wilson (3rd round) and Flynn (7th round) fall into this group it is the reason for all this, so on with the numbers…
The largest sample size is 9 for the All QBs Drafted 2nd or Below(QBs D2orB), starting in the league as of week 5 section and their career WP is just slightly above the average (0.92% not enough to change from a 9-7 record), but the All QBs D2orB since acquired WP is a 3% increase or a 10-6 record. Now that should be good news for Matt right? Well here’s where it gets tricky because the sample size drops down to three.
The nine QBs D2orB are divided exactly by three into the three categories; All QBs D2orB “Traded”, All QBs D2orB “Same Team”, All QBs D2oB “Free Agents” (FR/AG).
The All QBs D2oB Traded section did the poorest, but let’s forgets about that since it doesn’t involve our QBs.
The All QBs D2oB FR/AG, Flynn’s component, also did poorly. Not as bad as the traded QB’s, but enough to factor into a game loss or a lesser advantage over our base line for their estimated records.
Career winning percentage -4.8% barely enough for a 9-7 record [8.57-7.43]…
Since Acquired winning percentage -2.99% a closer to a 9-7 record [8.86-7.14]…
First year starter winning percentage -9.67% for an 8-8 record [7.79-8.21]…
And his next year winning percentage -4.06% for again barely a 9-7 record [8.69-7.31]…
Humm, not so good for Matt now.
On the other hand, the All QBs D2orB by the Same Team section has a significant advantage over the average base line, and that’s Wilson’s component.
Career winning percentage 16.78% for a solid 12-4 record [12.02-3.98]…
The Since Acquired winning percentage would remain the same a 12-4 record [12.02-3.98]…
First year starter winning percentage 7.36% for barely a 11-5 record [10.51-5.49]…
His next year winning percentage falls -1.21% not as much as the Free Agent category, but enough to drop the estimated record to 9-7 [9.14-6.86]
So what have we learned today?
1]Wow, what a breakdown I’m really glad I read all of it and now know more than I did yesterday.
2]What a waste of time I can’t believe I read the whole thing or even reading these stupid questions.
3] The author has way too much time on his hands and/or love for Excel.
4]The author uses statistics as a drunken man uses lampposts - for support rather than for illumination and should drink another beer.
5]The author should just go away and never post again.
Please feel free to take my survey or add your own answers….