I was just looking at FRCLocks, and I’m very puzzled by their numbers. It shows NRG 948 as only 73% chance, but we’ve already received the official email saying that we’re in. (70 district points; earlier estimate was that it will take roughly 61 to qualify. One district event remaining to play.) Teams with 65 points it’s showing as 25% and 20%.
Anyone know how FRCLocks is doing their math?
It’s right twice a day at least.
If you click the percentage for your team, the following page walks through the calculation.
– A similarly confused mentor new to districts and FRC locks
FRC Locks doesn’t use a percentage chance. The lock % is a weird metric. Basically, it’s the number of points the teams below you need to pass you and stop you from qualifying, divided by the total number of points remaining across the district.
It doesn’t account for probability, teams playing 3rd events, etc. Once you hit 100%, you’re locked in. Anything less than that, and the naieve calculation assumes there is some small chance you don’t qualify.
Hello yes this is my pet project. Thanks for the tag Andrew.
The locks are just “are there enough points left to kick you out if distributed to other teams in the worst case scenario” which can include strange edge scenarios that may or may not be realistic. The percent is not a prediction, just how close you are to crossing this threshold, a progress bar almost.
You say you have already received and invite for your district championship. Congrats! The folks who run your district are most likely looking at the historical points cutoff for your district and have determined that you will be attending based on that, which I would agree with.
The purpose of the website is, realistically, easy access rankings with quick visual cues showing if a team is in or not, but it really shines in the last week of district events as the last few points remaining get distributed and you watch the cutoff close in real time as elim matches happen.
If I had to guess, you would be shown as 100% and locked at the conclusion of qualification matches on sunday.
Any other questions I’d be happy to answer!
EDIT: oh yeah also, as someone said above, click on the % and I show my work.
https://frclocks.com/index.php?t=frc948
If the teams listed in the table earn the points listed, they would all tie with you and potentially knock you out depending on tiebreakers. If that points distribution seems unrealistic, you’re essentially a lock.
Given that explanation, there still seems to be something odd going on when teams have the same district points. It’s only looking at points to tie, not at tiebreakers, so why are teams with the same district points sometimes showing very different numbers?
For example: Teams 753 and 6443 both have 65 district points, but it lists them with different percentages. Looking at the expansion, 753 says 239 points needed to tie, and the list includes 35 points for team 4469. 6443 says only 204 points needed to tie, and has the same list of teams except that it’s missing team 4469.
(Of course, this analysis completely ignores whether it’s even theoretically possible to get that point distribution, given the discrete chunks that playoff and award points come in, but it sounds like that’s not the purpose of the site anyway. 
)
The rankings are using tiebreakers, team 753 has a higher tiebreaker than 6443. Because of that, team 753 needs ten teams to pass them to be eliminated, but 6443 needs only nine teams to pass them to be eliminated.
We calculate “points to tie” and assume that the tiebreakers for the new teams will push them over the team in question. Does that make sense? It’s asking “if the worse case scenario occurred, would this team still qualify for the district championship?”
With all this said I do agree the numbers are weird and not intuitive
This is where I would like the site to be eventually. Long term goals. Its not as easy as it sounds, but if you want to give it a crack my inbox is open to suggestions
Would it be possible to make it so you can see the points to tie breakdown for teams that win the Impact Award like every other team has?
I’ve really enjoyed this site for the last few years to track rankings and district point tallies. This is the first time in a while we’ve been in this position, but we also did well in points and would like to know where we stand in relation to our friends quickly.
Some years I’ve done a projection by giving every team that’s only played one event the same score for their second event, and teams that haven’t played at all the average of all teams that have. Not very realistic, but it might make an interesting comparison.
Unfortunately the way the math works, impact teams are essentially ignored and skipped over. But I can see your team has 106 points, and at this point you are a 100% lock anyway. Had you not won impact you’d have 96 points, and still be at 100%
Hmm, did I find a small bug?
Looking at NE, team 5563 has 36 points and is listed a 0% locked. The expansion says that other teams need 0 points (total) to tie 5563 and push them out, and lists 3 teams.
But those teams do need to get 4 points each. Those points are guaranteed if they just show up to their 2nd comp, but I assume that is actually 12 points. So shouldn’t the percent be 12 / 1748 = 0.69% ?
I guess maybe you are pulling the minimum 4 points for each teams out? Does that mean the “1748” does not include 4 pts/team who still have to play?
These guaranteed points are already considered in the calculations. Teams with a 0 for an unplayed event are assumed to have 4 points*. Thats why the “Adj. Points” column shows those teams having 36 instead of 32. These points are also removed from the points remaining pool for each event, as they aren’t actually up for grabs for the teams, we already know where they’re going.
*they actually are assumed to have whatever the last place qual rank points would be, as calculated by the monster formula in the manual. For bigger events (dcmp, usually), it actually ends up as 3 points
Thanks, we weren’t too worried about that. Just looking for a quick way to check bragging rights based on points against more than a handful of teams.
Appreciate the site!
