2020 RULES: Playoff Simulation Efficiency and Automation

[jmdaz44]

I'm kind of torn because I definitely like being able to capitalize on a player's big real game update and agree it would be easier to use that real game update across all leagues for that player. However, I like having the floor that a multi-game average provides. I get that I'm wanting the best of both worlds.

[Goodell]

Following up on this off-season rules discussion that impacts the playoffs. I put in a rules change suggestion widely supported that'll make managing the playoffs across multiple leagues easier and allow them to be done more efficiently.

If you click a skill position player's name on your team roster, you'll find their multi-game average on the bottom of their player page. We use multi-game averages so the results more fairly generate reasonable results for a player's average. If a QB averaged .4 interceptions per game, for example, we don't have them throw no picks through 4 playoff games unrealistically, but instead it's more like an interception every other game for that average.

In the past a team might have different players on their roster on different games of their multi-game average and made it very time-consuming and detail-oriented to sim. That might happen if some of the players used their average always but other players on the roster used a real update. I had to track it like okay the QB is on their 2nd game, but the RB is on their 1st game, this player is on his 3rd game average in one league but 2nd game average in a different league, etc. It was very time-consuming for me to track, and only more so as we add more leagues.

Instead of listing a players multi-game average for their 1st, 2nd, 3rd and 4th game personally as we did before, we instead now just list their multi-game averaged to be used universally in the first, second, third and 4th round of the playoffs. So all players not using a real playoff update will all be using the same round of their multi-game average. In the divisional round of the playoffs, for example, all players not using real game stats will all use their divisional round average across all leagues.

I also proposed a change to how we do the request to use a real playoff game. If I was going to automate things, I needed less manual changes and more universal data. So we were going to do it where if a player's real playoff stats were obviously better than their average for that round, that we'd use the real playoff game update for everyone in all leagues.

I haven't quite got things where I'd hoped with automation when proposing those changes. So since I'm still running the sims manually, we can still allow teams more freedom to choose to use a real playoff game or not. If it was automated, I needed universal data across all leagues and couldn't have a player with one update in one league and a different update in a different league. But since we're not there yet, I can manually use one update in one league and a different one in a different league if one GM says they want to use a playoff update and the other in another league doesn't.

So we'll implement part of the approved changes to make playoff simulations more efficient. Everyone will use the same multi-game average that relates to the particular round of the playoffs that week for efficiency and less individual player tracking, but we'll still allow teams to choose to use a playoff game or not without it having to be the same across all leagues since we aren't fully automated with the sim yet.

[bpboguta1483]

I like the real game playoff option just because GM's would need to analyze if their players would fare better in a real game or go with averages. For example the bears play the saints in the divisional round, would Anthony Miller fare better with game averages or the real game scenerio? Plus have a deadline in which the decisions are made, it's too easy to say use real or averages after the fact.

[Goodell]

Plus have a deadline in which the decisions are made, it's too easy to say use real or averages after the fact.

Just on that point, we have the deadline after the games purposely. Because playoff participants face significant injury risk compared to non-playoff players who can no longer miss games due to injury, having playoff players could be a severe DISADVANTAGE. That doesn't make sense, though, to penalize teams that have winning players on winning teams in reality. So because they have a disadvantage, we also give them a balancing ADVANTAGE that if they have a great performance in a real playoff game that you can benefit from that. You could suffer if they get hurt in real playoff game, but you could massively benefit from a strong playoff run in reality.

[jerrydlux]

In regards to the Derrick Henry 200 yd game example. I think if the GM didnt designate that they wanted real life stats then we shouldnt assume real life stats for that team. Yes, it is an obvious choice n that situation but I believe we also desire active GMs and part of that is being around for the playoffs to make that designation.

I also agree that the average stats lining up with the weeks of the playoffs makes sense. But will those averages change if players have a few big games during the playoffs? Yes, we would most likely use those real life stats for those games but what if they dont make the real Super Bowl, do we get a better average or drop back to the regular season average?

[Goodell]

But will those averages change if players have a few big games during the playoffs?

We don't recalculate the seasonal average every week of the playoffs. It's just a regular season average with fixed numbers from week 1-17 regular season real NFL games.