[FF]: 2017 Fantasy FIRST Proposed Rule Changes

[tindleroot]

Created for the discussion of any proposed rule additions, subtractions, or modifications for the 2017 FF season.

[tindleroot]

Recently I've been thinking about the waiver system - waivers and free picks have started to evolve into an "arms race" situation where the faster person gets great teams at no cost to themselves, certainly creating an unfair scenario for teams like RFA, FP, and TBC who all took advantage of the system. So I started looking at all of the reasons a team might want to swap:

1) A team they picked is no longer attending the event, so they opt to replace them.
2) A good team has recently signed up for the event.
3) One of their picks played very poorly at a previous event in the season.
4) An unpicked team played extremely well at a previous event in the season.
5) An picked team earned a lot of points at their first district event, making a high average.
6) One of their picks played poorly at their first district, lowering the average.

Given the reasons above, I think having a more-inclusive waiver system is in need for every case listed above except (1). If a team dropped an event, then the FF team gets first dibs at another team, regardless of waiver order or date (perhaps this should close by the start of the event?)
Otherwise, if anyone wants to swap a team for reasons 2-5, a waiver use should be required, and the team they want to swap needs to be posted publicly and have 1-2 days in which anyone else can take claim. As for districts, I actually support the 1-play swap idea, but in addition to waiver we should set a maximum on the number of district swaps allowed, perhaps around 3-5. This means that you have to judge swaps carefully, and you can't expect to get a lot of unexpectedly good teams, rather you have to put more thought into the original picks.

Another thought I had for the waiver system was to create an auction-based system. Give ever FF team x waiver points to start out with. If you wanted to claim a team, you would do so with an attached price, but someone else could out-claim you, even if you were ahead of them in the order. In this way, an FF team could either claim a few really good teams, or a lot of decent teams, or a mix of the two, making for a fairer waiver process.

On to another topic, inter-district plays.

This one is going to have to be a compromise in some way. If we count inter-district plays for points, then averages could be reduced since teams are ineligible for the legacy awards at those events. If we don't count them, then inter-district teams will have a higher average than most since they will have a practice event that doesn't count for points (just like FRC). I really want to lean towards counting those teams for points at all events, but something has to be put in place if this happens in order to account for awards.

Just some initial thoughts.

[jlmcmchl]

Originally Posted by tindleroot

This one is going to have to be a compromise in some way. If we count inter-district plays for points, then averages could be reduced since teams are ineligible for the legacy awards at those events.If we don't count them, then inter-district teams will have a higher average than most since they will have a practice event that doesn't count for points (just like FRC). I really want to lean towards counting those teams for points at all events, but something has to be put in place if this happens in order to account for awards.

This has been a big point of discussion this season, and I'll link my post that starts the discussion in the main thread.

http://www.chiefdelphi.com/forums/sh...&postcount=282

I'm curious as to why you think teams should necessarily get credit for inter-district events. As you said, their 'point pool' is significantly limited, compared to teams in that district. Would you draft a team who is going to be handicapped like that? I don't think I would.

This might be fixed, like you said, by an overhaul of the district points system.



Another discussion I would like to have, since this has now come up, is the points system for scoring districts. Some teams, like 68, go to 3 or more districts. Food for thought:

Team A is a decent team, places about 15th and goes out in quarters in their events. They're a community driven team, though, and are expected to win CA each year. Team A intends to go to 3 district events this year.

Team B is from the town over, and is about as good on the field as Team A. It's well organized and makes a big difference for raising rookies in the state. As such, they are expected to win EI this year. Team B is scheduled to go to district events.

Team C is from across the state, and also is middle of the pack on the field. However, they put a lot of pride in their robot, and are expected to get a robot award at each competition. They will attend 2 district events.

Their FF points look as follows:

Team A:
((<Qualification Score> + <Seeding Score> + <Playoff Score>)x3 + 42 (Chairman's)) / 3

= <Qualification Score> + <Seeding Score> + <Playoff Score> + 14

Team B:

((<Qualification Score> + <Seeding Score> + <Playoff Score>)x2 + 36 (Engineering Inspiration)) / 2

= <Qualification Score> + <Seeding Score> + <Playoff Score> + 18

Team C:

((<Qualification Score> + <Seeding Score> + <Playoff Score>)x2 + 15(Robot)) / 2

= <Qualification Score> + <Seeding Score> + <Playoff Score> + 15


It doesn't make sense to me that Team B gets more points than Team A in this situation, strictly as a result of going to one less event. Team C was provided for reference, as it does take a lot of effort, and a little luck, to get two robot awards in a year.

I propose this change to the district points system:
Field performance is normalized independently of Award points. It can be normalized to one or 2 or however many events, but it should be constant across teams.

Awards are normalized to 2 events. So, in this case, Team A would be awarded 21 points for awards, over Team B and Team C.

This can be rephrased or reinterpreted through discussion. One option to implement this is to only consider awards acquired within their district, and normalized across those events. The more I think about it, the more I like it.

[ttldomination]

Originally Posted by tindleroot

...the team they want to swap needs to be posted publicly and have 1-2 days in which anyone else can take claim...

I agree with you on everything except for this.

I agree that the current free agent process has become a big of an arms race, but we (RFA) work hard to identify swaps. As such, all that hard work becomes marginalized if other FF teams are free to just wait around and snipe waivers.

With that being said, I would recommend that waivers be handled 100% privately. The only thing that everyone gets notified of is the fact that a FF team has been granted a waiver.

Originally Posted by tindleroot

Another thought I had for the waiver system was to create an auction-based system. Give ever FF team x waiver points to start out with. If you wanted to claim a team, you would do so with an attached price, but someone else could out-claim you, even if you were ahead of them in the order. In this way, an FF team could either claim a few really good teams, or a lot of decent teams, or a mix of the two, making for a fairer waiver process.

While I suppose this addresses the situation of an 'arms race', I feel like this system will be a downright cluster to implement.

Originally Posted by jlmcmchl

I propose this change to the district points system:
Field performance is normalized independently of Award points. It can be normalized to one or 2 or however many events, but it should be constant across teams.

Awards are normalized to 2 events. So, in this case, Team A would be awarded 21 points for awards, over Team B and Team C.

What's interesting is that in your hypothetical scenario, Team A is penalized for attending 3 events. However, in your solution, a much more likely realistic scenario is that Team A is being awarded for attending three events.

I feel like your proposed scenario is much more black and white than reality would suggest. That is to say that Team A is not just going to win DCA; they're in the running for EI, a handful of the other non-engineering awards, and even the engineering awards.

Normalizing 3-event attempts worth of awards down to 2 events is ignoring the fact that Team A had three events in which they had the opportunity to win awards, which pits Teams B and Team C at a disadvantage.

- Sunny G.

[BrennanB]

Okay, not sure how many thoughts I will be able to hash out in one sitting, and i'm sure I will have more

#1 - Newly added teams should never be considered free agents. The only free agents are the ones that were on the draft originally.

Leads to a silly arms race, and not in the "we found them and you didn't" way, which it think is totally legitimate. At TBC almost every single RFA waiver, TBC was aware and wanting to take that team, yet they just posted first, cause I forget/don't stay up till midnight/have school stuff to take care of. This kind of "I posted first so I get the team" is lame. I think a waivers or nothing system is way more fair. This should probably extend to all second play teams, as the exact same thing happens. A smart team like TBC or RFA won't post in the waivers so that their team can get sniped, and will just snipe it once they are a "free agent" so lets do away with that. Lists lock on Tuesday. Gives you two days to process any last minute waiver claim, and if you want to add a team after that? Well sucks.

#2 - We need to fix draft order to make it more fair.

Really simple fix that is already miles better than our current system, though can be improved on:

Draft One:

Team One
Team Two
Team Three
Team Four

Draft Two

Team Four
Team One
Team Two
Team Three

Draft Three

Team Three
Team Four
Team One
Team Two

etcetc.

#3 - Out of district teams don't count for FF points since you can't win as many awards.

#4 - Waivers held privately

I like the idea of giving reward to the teams that find and spend the effort looking for good teams. Its part of the reason that the whole free agents arms race is happening. Though doing this privately is a little hard, since one person will always know what each team is requesting.

[tindleroot]

Originally Posted by BrennanB

#4 - Waivers held privately

I like the idea of giving reward to the teams that find and spend the effort looking for good teams. Its part of the reason that the whole free agents arms race is happening. Though doing this privately is a little hard, since one person will always know what each team is requesting.

What if we had a google form or the like where you can submit any waivers you want? Waivers close every week at a set time (Say Tuesday, midnight). Then the sheet is checked and waivers are assigned.

[EricH]

Waiver Item #7 (and the PRIMARY reason for waivers, originally):

Somebody bites off more than they can chew and drops or is booted from the league. Now there are a bunch of (probably-decent) teams available.

Also, Case #1 on waiver is the current system. Missing team? Grab a new one, you've got priority, no matter if the team is on waiver or not.

Looking at cases #2-#4, I would say that Case #3 isn't necessarily a waiver--they just want to drop the team for another team. BUT, if it's combined with #2 or #4, that's definitely a good argument for a waiver.

I'd be on-board with a Google form/sheet/other similar device for putting in waiver claims. What I'm less confident about would be full privacy. I think announcing that team X has had a claim made--and, once regionals start, all claims that aren't for a missing team are automatically waivers--should be enough. If somebody was sneaky about it, there might be a way to funnel the google form to a (private-ish) google sheet that would sort out a list of teams on waivers (but not who requested them) and put that on another (public) google sheet... At one point I was going to try to get a waiver-processing spreadsheet working, too. The logic is pretty crazy, though (check for all claims on this team just awarded and make them not matter--it's worse if there's a tiered draft, which we didn't have this year). Might try that as a summer project...



Brennan, regarding the draft order, I believe RFA members and others have posted in the main thread showing that everybody drafted at approximately the same position, on average. I think there might have been one or two teams--max--drafting significantly (minimum one spot) below the normal area. I'd say that that's one option, but it could be significantly flawed if the # of drafts and the # of players didn't match up fairly closely. For District drafts, that'd probably be the better option: Randomize the first one, and rotate the rest based on that. (Once all areas are District, then of course it'll be a moot point.) I was going to pull that for the DCMPs, but was reminded that those are by District score.

[BrennanB]

Originally Posted by EricH

Brennan, regarding the draft order, I believe RFA members and others have posted in the main thread showing that everybody drafted at approximately the same position, on average. I think there might have been one or two teams--max--drafting significantly (minimum one spot) below the normal area. I'd say that that's one option, but it could be significantly flawed if the # of drafts and the # of players didn't match up fairly closely. For District drafts, that'd probably be the better option: Randomize the first one, and rotate the rest based on that. (Once all areas are District, then of course it'll be a moot point.) I was going to pull that for the DCMPs, but was reminded that those are by District score.

Yeah while the average may be "closeish" the first/last picks were terrible. If you look at the points distribution over first/last picks quite certain the top 2 picks will on average have a much much higher point average than any other position. So just because your "average" draft order may be semi acceptable, the huge disproportionate first picks between teams I think is a big big issue. You can see that in my post on the main thread.

[EricH]

Originally Posted by BrennanB

Yeah while the average may be "closeish" the first/last picks were terrible. If you look at the points distribution over first/last picks quite certain the top 2 picks will on average have a much much higher point average than any other position. So just because your "average" draft order may be semi acceptable, the huge disproportionate first picks between teams I think is a big big issue. You can see that in my post on the main thread.

Ah, yes.

So the real question is not, how do we keep the draft order at acceptable averages.

It's "How do we keep the top 2 (or 3, if it's a deep event) spread evenly so that everybody has a crack at those top points?"

And that could be accomplished by a rotation, but then there'll be that one player who doesn't get as many if the number of events and players doesn't line up right...Or you get folks complaining that so-and-so always gets the top picks at the good events (2056 at just about any event, streak or no streak, just to name one).

What might be done (this is notional) is to randomize as normal, BUT if certain parameters aren't met, re-run the randomization. That is, if team X isn't getting good positioning, then re-randomize (or non-randomize) them into good position until it evens out a bit. Or if team Y is getting really good positioning, same but into a "worse" position.

I do recall that there was one year where the randomizer (not the spreadsheet in this case) just really liked one person... to the point where I may or may not have re-randomized if they came up near the top of the list...

[BrennanB]

I suppose we should bring this stuff back up as it's almost time to begin signups.

[Rangel]

Never even saw this thread. I am totally in favor of their being a easier to see interface in terms of waivers like a google spreadsheet. I'd be interested in playing the waiver game but it's hard to keep track of when claims are scattered across the giant main thread. A google sheet showing due dates in would be much appreciated. I'd be interested in putting that together if it's not already being worked on.

Also I wouldn't mind the picking order being modified to make positions 1 and 2 less overwhelming. Something like this I would like:
1-8
8-1
8-1

It just doesn't make sense to me why the first pick doesn't get the last pick except for maybe the championships draft where the top pick slot was earned.

Also I agree about the rotation for draft order. I 100% support that idea.

[EricH]

I'll be putting out the official thread before long, but here are the current changes:

1) There will be a rotation, which will re-randomize every round. (Basically, run through with one random order, rotating every draft, and then run a different random order and rotate that when the first one is done.) One note--and we'll take a look at this later--is the "remnant" drafts where there aren't enough for a full rotation. Those will either be full random or another rotation. District drafts will have their own rotation if I can foot that.

2) After events begin, all teams being picked up will be on waiver. (It's easier to keep track of that than of who's already played.)


Waiver transparency: I'll be working in free time on a waiver processing sheet, and will post it to CD-Media when finished (or at least workable). Some people like to not have others know who they are, for some reason. Input may be a hair interesting. Also, I think a dedicated waiver thread will be in order--that's something we've done in the past, and didn't manage to do this past year.

[BrennanB]

I guess the next thing to bring up is do we want the waivers to be public? My vote would be no, and I seem to recall RFA having a similar opinion. Having them private allows for that fun competitve edge on finding important teams to swap in, without the "whoever posts first" side of it.

Update waivers weekly on say... Tuesdays, and have them submitted via gdoc forum?

[Rangel]

Originally Posted by BrennanB

I guess the next thing to bring up is do we want the waivers to be public? My vote would be no, and I seem to recall RFA having a similar opinion. Having them private allows for that fun competitve edge on finding important teams to swap in, without the "whoever posts first" side of it.

Update waivers weekly on say... Tuesdays, and have them submitted via gdoc forum?

I really like the idea of a set day for waivers to be submitted. Would make it easier for many to participate in waiver wars since it can be scheduled on when to check and submit them. This is especially critical during the middle of build season when it's much harder to make time for FF.

[EricH]

Originally Posted by Rangel(kf7fdb)

I really like the idea of a set day for waivers to be submitted. Would make it easier for many to participate in waiver wars since it can be scheduled on when to check and submit them. This is especially critical during the middle of build season when it's much harder to make time for FF.

Or during competition season.

I'm going to leave the discussion open on waivers for a while, as I'm sure there are more items to tease out to improve that side.


The other discussion item I'd like to note: there aren't a lot of player-to-player trades in FF. They're legal to make (team or draft slot or some combination), just wondering if there was any way to incentivize those. I don't think I've seen one in 3-4 years.