RealGM Top 100 All-Time (2017)

[dhsilv2]

As far as a ballot/point system (e.g. 1st ballot choice worth 3 pts, 2nd ballot worth 2 pts, 3rd ballot worth 1 pt.....or similar point system, where winner is decided by who has the most points): we've intentionally avoided that type of system as it provides the GREATEST ability to "manipulate" the vote against undesired candidates.

However, an idea that your suggestion brought to mind (and maybe this is what you meant) which I potentially would have been open to is having participants specify an "Alternate #2" [i.e. 3rd choice] vote which would get transferred ONLY in the event that their Alternate #1 becomes a ghost vote. And if that Alternate #2 pick also ends up going ghost, then that's that.

idk though, I'm reluctant to change the protocol in that way here at the halfway point of the project unless there's a lot of support for the idea.
It's definitely food for thought for future projects, though.

Might be good to do, especially with how open these portions of the ranking are. The amount of ghost votes will definitely be increasing a lot otherwise.

Thoughts?

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Full honesty and not to be negative, but with how scattered the votes are getting and how few people are voting (the early rounds were getting 12+ pages). I'd be ok with a run off on the top 3 unless we had a clear choice from here on.

The problem for me is if I vote early, I can't "adjust my vote" as easily. I'd rather we add a day to this and do a run off vote for the top 3. I know that adds some effort, but unless we can get 30-40 votes I'd rather we pick between the top 2-3 (I say 3 so we can do 2 point 1 point thing just to make it clear). Plus I think we get better picks.

Adding a 3rd alt doesn't help much, I think at this point there are 30 players who we can consider so we'll be all over the board.

[Dr Positivity]

I'm in support of the run-off system as well in the case of a non-majority

[Doctor MJ]

Honestly, I think a run-off is the best way to do this. Run-off's give the extra work an excitement and remove the need to feel committed to a position prematurely.

But really, I don't think it's that big of a deal either way. Whatever you decide, it's fine.

[euroleague]

I suggested this idea a few threads ago.

It's pretty rare someone convinces me to change my vote at this stage, since I have Cousy ranked pretty high. Normally I plan my votes several threads in advance, seeing premature discussions for players who are suggested. I think 3 votes is better to avoid ghost votes going to nothing, and 3 is a good number in general.

[Winsome Gerbil]

I was thinking we could maybe switch to a simpler and more logical system.

1) a new thread is opened up
2) I post who I think should be the next guy
3) everyone else posts to say they agree with me

after we get the timing streamlined, we could probably even do multiple threads a day that way.

[Outside]

Main downside of the double alternate for me (not for everyone I know) is that I am still making up my mind by reading other people's ideas and posts and comparing them to my own preconceptions. It forces me to decide farther in advance and then I tend to be defending my choices rather than making new decisions.

I agree with this argument against multiple alternates. If I feel strongly enough about a third guy, I list him as an honorable mention so as to introduce him into poster's minds for a thread or two down the line, and that's good enough.

While the efficiency of Winsome Gerbil's system is undeniable (thanks for getting my day started with a smile), a runoff makes the most sense to me.

[Hornet Mania]

I'm also in favor of a runoff system.

[Pablo Novi]

Your suggested tie-breaking procedure sounds very good to me (and adds to how impressed I've been with both the general procedural guide-lines AND actual care of carrying them out).

A thought / question:
If we had people vote for THREE instead of TWO players (Vote, Alternate #1, Alternate #2); that should help in terms of reducing early-tallying ghost votes; but wouldn't it also better reflect people's preferences?

As far as a ballot/point system (e.g. 1st ballot choice worth 3 pts, 2nd ballot worth 2 pts, 3rd ballot worth 1 pt.....or similar point system, where winner is decided by who has the most points): we've intentionally avoided that type of system as it provides the GREATEST ability to "manipulate" the vote against undesired candidates.

However, an idea that your suggestion brought to mind (and maybe this is what you meant) which I potentially would have been open to is having participants specify an "Alternate #2" [i.e. 3rd choice] vote which would get transferred ONLY in the event that their Alternate #1 becomes a ghost vote. And if that Alternate #2 pick also ends up going ghost, then that's that.

idk though, I'm reluctant to change the protocol in that way here at the halfway point of the project unless there's a lot of support for the idea.
It's definitely food for thought for future projects, though.

My suggestion was not for a 3-votes system (recognizing the increased possibility / potential for result-manipulation); but, instead, "(and maybe this is what you meant)" "Alternate #2 [i.e. 3rd choice] vote which would get transferred ONLY ... ghost vote." etc.

In other words, what I meant was to propose a small adjustment to help reduce early-tally ghost votes while hopefully producing overall results that better reflect our thinking.

As to changing the protocol in mid-course - I wasn't actually calling for that; though I certainly wouldn't be opposed.

[Pablo Novi]

I was thinking we could maybe switch to a simpler and more logical system.

1) a new thread is opened up
2) I post who I think should be the next guy
3) everyone else posts to say they agree with me

after we get the timing streamlined, we could probably even do multiple threads a day that way.

Sorry, and it could just be me, but I have no idea what you are proposing here.

[Outside]

I was thinking we could maybe switch to a simpler and more logical system.

1) a new thread is opened up
2) I post who I think should be the next guy
3) everyone else posts to say they agree with me

after we get the timing streamlined, we could probably even do multiple threads a day that way.

Sorry, and it could just be me, but I have no idea what you are proposing here.

I think you missed the joke. I thought it was funny, but then I've got a different sense of humor.

[Pablo Novi]

I was thinking we could maybe switch to a simpler and more logical system.

1) a new thread is opened up
2) I post who I think should be the next guy
3) everyone else posts to say they agree with me

after we get the timing streamlined, we could probably even do multiple threads a day that way.

Sorry, and it could just be me, but I have no idea what you are proposing here.

I think you missed the joke. I thought it was funny, but then I've got a different sense of humor.

Dang!
Somebody like me with a very different sense of humor and .... and
I didn't get it. (At least I didn't flame war or something - you gotta give me that, right?)
lol

[Pablo Novi]

Regardless of the change ("2nd Alternate" or a run-off), I THINK (but what do I know?) that it would be worth it to try it now / soon in THESE GOAT threads - so that if there's a problem we-all find out now (when the individual rankings of players is not nearly as controversial as higher up the ranks) - not only for this 2017 vote but also for future ones. And if there's no problem - then that's good too for both now and the future ones.

[trex_8063]

OK, there was only lukewarm support for the idea of a 3rd pick ("Alternate #2") idea, so I'm scrapping that. Several people (who have posted here or contacted me via PM) appear to be in favor of some manner of run-off as needed.
However, I'm reluctant to utterly scrap the protocol we've been following, and also still have an interest in keeping things moving (as project fatigue has already become a significant issue). So I'm adopting the following modifications, which I have edited into the OP:

EDIT (updated 9/7/17): In the event that this procedure narrows it down to two individuals who have the same number of votes, we will at that time enter a 24-hour runoff vote in which those participants who had a counted vote in that thread and who did NOT cast either their primary or secondary vote for one of the two finalists will be asked to pick between the two finalists as a means breaking the tie.
In the event that this is still insufficient for a tie-breaker, OTHER registered participants who did NOT have a counted in that thread will be solicited to pick between the two finalists (poll open for ~24 hours) as a means of breaking the tie.
Beginning in thread #40 and henceforth, though we will continue to use the ranked choice system we have been to this point, in any thread in which there are fewer than 15 voting participants we will AUTOMATICALLY enter a run-off [same guidelines as outlined above] between the top two [*or 3-4, as the case may be in some instances] vote recipients if there is no true majority, even if one candidate has a lead in the vote.
*If a run-off between three or more candidates results in a tie for 1st, the 3rd [+/- 4th] place finisher(s) will be eliminated and anyone who voted for an eliminated candidate will be asked to choose between the remaining two candidates as a means of settling the tie. Similar procedure will be followed to "whittle down" the candidates as needed in the event of other ties in multi-player run-offs.

I think we've technically already violated the "rule of 15 voters" once already (I recall at least one thread in which there were only 14 votes). However, scant vote totals are more problematic here than they were at that time (and it only gets worse). Once out to #40 (and even more so in another 20-40 places), the field is just so wide open.......we may well see a thread in which we have like 11 votes: 2 votes for Player A, and 1 vote each for nine other individuals (and in which many of the alternate picks will become ghost votes).
So everybody continues to pick a primary choice and an alternate, but from here on, in any thread in which there are not at least 15 participants, we automatically have a run-off if there is no majority. And run-off will otherwise be utilized to settle any other ties.
We'll see how this goes for a few more threads. It's possible additional modifications will need to be made in the future. While the field gets more and more wide open the further out we go, sadly participation tends to dwindle more and more, too.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

[euroleague]

...

There is another scenario. Such as, if it comes down to 10 individuals all tied with 1 vote. Or, five individuals with 2 votes. Which seems quite likely.

I think "Alt" votes should count for half a point, and be used to decide ties such as this, where we include everybody's alt vote. (even if their vote is still in the running)

[trex_8063]

fyi, I made one adjustment to the run-off procedure which is that it will NOT initially be limited to ONLY those participants who'd cast a vote in that thread; it will be open to ALL registered project participants.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

[trex_8063]

...

There is another scenario. Such as, if it comes down to 10 individuals all tied with 1 vote. Or, five individuals with 2 votes. Which seems quite likely.

I think "Alt" votes should count for half a point, and be used to decide ties such as this, where we include everybody's alt vote. (even if their vote is still in the running)

That's basically a ballot point system, which we've intentionally avoided as it allows the greatest ability to manipulate the vote.

[THKNKG]

...

There is another scenario. Such as, if it comes down to 10 individuals all tied with 1 vote. Or, five individuals with 2 votes. Which seems quite likely.

I think "Alt" votes should count for half a point, and be used to decide ties such as this, where we include everybody's alt vote. (even if their vote is still in the running)

That's basically a ballot point system, which we've intentionally avoided as it allows the greatest ability to manipulate the vote.

Thanks for running this thing so well.

[euroleague]

That's basically a ballot point system, which we've intentionally avoided as it allows the greatest ability to manipulate the vote.

I think many are already changing their votes based off who is popular.

Just use the honor system. If you are influencing the vote it means you're lying about your picks - which can make you look dumb in the long run.

[trex_8063]

That's basically a ballot point system, which we've intentionally avoided as it allows the greatest ability to manipulate the vote.

I think many are already changing their votes based off who is popular.

Just use the honor system. If you are influencing the vote it means you're lying about your picks - which can make you look dumb in the long run.

I've stated previously I'm fine if people go 2-3 spots further down on their list [particularly if talking about their alternate vote] to someone who has a little traction, if the person you would otherwise choose has none. I leave that to individuals as to how they want to treat that.

Is that "strategic manipulation"? idk, it's perhaps walking the line a little. But a true attempt at manipulation would be more along these lines:
Suppose Players A & B are my ideal 1st and 2nd choices, respectively.....but neither has any real traction presently. Meanwhile, Player C----who I really think is overrated and doesn't belong anywhere in the next 15 spots----is one of the top contenders to take the spot.

Players D & E are guys I have just 2-5 places below my ideal 1st choice (Player A) on my own ATL, and both are just starting to get a small amount of buzz/traction......but probably not enough to challenge Player C.

So I cast my votes for Players F & G, who I actually have between 6-10 places lower than Player A on my personal ATL (and thus a few places lower than D & E) because they're the two with the most potential to beat out Player C.


That's the kind of thinking that I would frown on.
And with the system we're currently using, my vote could ultimately only go for ONE of Players F or G.....but not both; and the value of that support is always the same (1 vote). In a point system, I can stack the deck for BOTH Players F & G, against my hated Player C (and I could further strategically decide which of Players F and G to give MORE points to, because 1st ballot is worth more than 2nd).

[trex_8063]

I had thought the recent changes I’d made to the voting protocol (“rule of 15”, run-offs with any tie, etc) would be sufficient to eliminate potentially spurious results. But the last thread has me thinking I was mistaken. Even with a relatively nice turnout of 21 voters, when that’s spread out over 14 candidates, it can still be a bit “flukey” if we use only the ranked choice system as the sole means of deciding a winner. And again, I expect things to get even worse the further out on the list we get. And multiple people have been grumbling about similar concerns.
So I’ve made the following alterations to the protocol, effective immediately:

EDIT (updated 9/9/17): Beginning in thread #41, we will continue using the ranked choice vote as we have been as a means of narrowing the field down to THREE candidates (or potentially 4-5, in the event of a tie for 3rd place, or a 3-way tie for 2nd, etc). We will then automatically enter a run-off [which will be open to ALL registered project participants, even if they had not cast an initial vote in that particular thread] between those 3+ candidates, where participants are asked to select which ONE of the finalists they favor (at least cursory reasoning for one's pick is required). In the event the run-off leads to a tie for 1st place, we will enter a second run-off between only those two individuals.

In other words, we’ll continue to specify a primary choice and an alternate, and I will tally and eliminate candidates based on the ranked choice system as I have previously…..but only to the point that we get down to the final 3 (or more, as it will sometimes be) candidates. And then we will enter a run-off vote between the finalists.

Very similar to how we did it in the 2014 project, except that with our alternate picks we each have a slightly higher probability of having played a part in the selection of the finalists. The only other difference being that I’m going to try the run-offs between 3(+) candidates instead of just the top 2.

This adds time to the project, but I don't see a way to better the "results" that doesn't lengthen the project. This hopefully will inject a little more discussion to each thread, too.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.