Steg! The first of the picks from the mega trade.
(03-16-2020, 09:33 PM)iStegosauruz Wrote:As I explained in methodology I inserted Farley Hank into y'alls depth chart as a starter in two formations and a backup in the other three. If you tell me how you intend to swap players around I can definitely run the simulation again and get you the numbers - or you can provide me your internal testing numbers although I'd still likely back check them. Well, I can tell you that unless you moved AE to CB, adjusted his attributes, updated the DC based off that move, and then changed our playbooks to optimize win %, you'll get worse results than what I tested. Unfortunately, I am not going to leak our season strats ahead of the game week occurring, and since I change our DC/playbook/run-blitz %/tempo weekly tailored to our individual opponent, I would say using base strats across multiple seasons of data is going to return you flawed data. (03-16-2020, 09:33 PM)iStegosauruz Wrote:As I also stated in the piece I did experiment with different variations - swapping Leaf to WR and TE in various spots/moving Farley around where it seemed like the advantageous thing to do. I didn't save any of the numbers because they were smaller sample sizes and it was more for margin of error testing, but it didn't have a substantially greater impact. I'd still, on whole, say you gave up immense value. Again, did you move our #1 WR to TE? Because adding Hank to the offense wont do much, but adding a 1000 TPE corner to the defense has more effect than a marginal change. (03-16-2020, 09:33 PM)iStegosauruz Wrote:Your top four TPE players are already regressing or will be after this season. I understand the idea that the window is now but your expected winning percentage before the trade was 37.08%. Its going to take a lot of big moves to turn that into an Ultimus contender - and the point is I don't think this trade was that type of move even with any potential margins for error in my calculations, especially considering the capital you gave up. Hence the article - based on a wholistic analysis of the situation you probably should have kept the draft picks and pivoted for the future. Hey, I appreciate the feedback. God knows my numbers and system isn't infallible, so seeing actual fact-based analysis, rather than the usual knee-jerk reaction-based response I normally see, is refreshing. But I still think the bulk of your analysis is built using flawed data.
Something not talked about, but I thought about - those 1st and 2nd rounders would be auto-protected if there’s expansion. This trade means that Chicago now has to either use two protection slots on these players, which takes away from two they could have protected before, or let them walk anyway.
With that being said, I can see why Chicago did it. The NSFC is wide open this season. Only Baltimore has a positive differential rn. If Chicago makes the playoffs, honestly, anything can happen, and that’s probably what they’re banking on. (03-17-2020, 09:34 AM)steelsound Wrote:Well, I can tell you that unless you moved AE to CB, adjusted his attributes, updated the DC based off that move, and then changed our playbooks to optimize win %, you'll get worse results than what I tested. Unfortunately, I am not going to leak our season strats ahead of the game week occurring, and since I change our DC/playbook/run-blitz %/tempo weekly tailored to our individual opponent, I would say using base strats across multiple seasons of data is going to return you flawed data. Yeah totally all fair - there is going to be a gap in results if you move him to CB but if you replace your second lowest CB you're getting a 757TPE increase at first CB. Second corner increases by 343TPE. At that point Leaf can't swap again from what I've been told so you either have to swap Farley to WR or you're running a 287TPE at your second WR and a 650TPE at your first WR. Thats a 337TPE reduction at first receiver and a 363TPE reduction at second receiver. Thats a 700TPE change at receiver for a 1100TE change at Corner. I can buy that all that realignment gives you a marginally better chance at success than before but I find it hard to believe that it makes up for the 37.08% control study to truly launch you into contention. I'll run the numbers on it with default WK1 playbooks later today or tomorrow though just to back check this. Its all theory and I understand you believing in the moves you made - I expect nothing less - but any way you slice it if even if it works out its still an extremely risky move, which was the underlying message of my post and my original reply to you. Its an innovating but RISKY approach to teambuilding that requires moving the needle for you substantially from where it was before to make a longterm impact. (03-17-2020, 09:21 AM)iStegosauruz Wrote:Yeah totally all fair - there is going to be a gap in results if you move him to CB but if you replace your second lowest CB you're getting a 757TPE increase at first CB. Second corner increases by 343TPE. At that point Leaf can't swap again from what I've been told so you either have to swap Farley to WR or you're running a 287TPE at your second WR and a 650TPE at your first WR. Thats a 337TPE reduction at first receiver and a 363TPE reduction at second receiver. Appreciate the reply! So a few more thoughts on this: 1) The offset TPE amount on offense is exactly the amount which we just traded for - ~750 TPE 2) Leaf can't swap but running 2 RBs allows one to be designated as FB, which allows a RB to play at outside WR. Now, they can't be designated primary WR but they can still take snaps there and carry as a RB in some sets. So there is some plan to do that (not really a leak there) 3) Using week 1 playbooks will give a different result against different teams, as tempo changes week to week can drastically change your win % depending on the team you are facing (for example, going from Very Fast to Normal changed ours 5% week 2 or 3, whenever that change was made). This can have a cascading effect if you don't set up your tests to optimize vs. each team. This is really good feedback though, and am loving the opportunity to cross-examine my processes and data. Keep it coming! (03-17-2020, 12:10 PM)steelsound Wrote:Appreciate the reply! So a few more thoughts on this: Yeah I get the acquisition TPE amount - makes sense. And I get the designation as FB playing outside WR. When I ran my Approximate Value study Irving was classified as a FB and got most of his usage through the air - good example of that occurring. And without a doubt situations are going to change as you scheme differently - some things I can't account and for and inevitably end up in the margin for error of whatever analysis I do. Obviously using pieces acquired in a trade in the most optimal way possible will increase their effects - optimal having some variance every week as well depending on scheme, matchup, etc. I'm going to revisit all of this when I have some time and run it with your position changes in mind and with optimization toward team sets, etc - inevitably we'll see the effects this has y'alls team later on anyway which will help me refine my takes on this stuff and ability to analyze them in the future. I'm starting with a very narrow base. Easier to refine methodology when I have a breadth for what actually works when put into practice. To plug another article I just did on containing opposing quarterbacks it looks like y'alls defense has been the let down - hence why moving AE to CB and bumping up your TPE allotted there makes sense. So in a vacuum that'll help out a glaring problem. |
|