Scoring improvement on the way!

#1

Hey everyone,

We just wanted to let you know that our points system itself is not changing, but we’re planning to make a little adjustment to the timing of when traces are scored. It should only increase the number of points you earn; here is just a quick breakdown of what’s going on.

Right now, we know everyone hates being the second player on a cube, tracing it correctly, and then receiving a low score because the trailblazer missed a bunch of stuff. This can be even more frustrating if your trace is still immediately added to the consensus, since your accuracy will update but you’ve been cheated on those points forever. The main reason such circumstances can arise is because the scoring program has been running before consensus updates with your trace. As part of smoothing out the discrepancies between what you see in the Activity Tracker, Review Mode, and the points you actually receive, Chris will soon adjust scoring to calculate your points after consensus updates with your trace. You will thus never miss points for adding stuff as the second player if the TBer missed it. It does open an opportunity for second players to overcolor and receive erroneous points, but since this is a less common problem, we hope it’s a fair trade.

TLDR: you will still receive points instantly as the second player (or third, or fourth…) player on a cube, but in terms of the split-second calculations and adjustments that happen in the database, second-players will now have a much higher likelihood of being fairly compensated for their traces, and poor TBs will have less effect. None of these changes will result in you getting less points. Only more. More. Mooooarrrr!

If you have any questions, let us know here or in chat!

3 Likes
#2

Really enjoying that news !

So, it will also fixe the tracker/review mode not following consensus correctly ?

/me thinking playing again… :stuck_out_tongue:

#3

This change ought to roll out as part of fixing those issues, yeah. My understanding from Chris is that right now it’s all an interconnected problem of certain pieces of the code referencing consensus at different times. For maximum consistency we can also adjust the scoring timing alongside the other stuff, and the benefit to players in that case would incidentally be what’s described above. We can’t really think of any cheating scenarios that could occur, because to intentionally overcolor for extra points as player 2, you’d have to know that’s who you were, which generally you don’t.

1 Like
#4

Actually…if tber misses part of trace and 2nd player adds it, spawner updating wt (and consensus from 1 to 2) won’t add the part of trace until and if 3rd player adds it. 1 v 1 doesn’t add, lol. 2 v 1 adds. 1 v 1 keeps trace out.

The improvement will help greatly when someone’s 3rd. meaning that tber missed, 2nd added, you add as 3rd.

As 3rd so far you wont get points but accu fixes as it’s added, with the improvement 3rd will also get points but 2nd won’t unless 2nd gets points after wt 3.

Unless something’s changed or will change that’ll invalidate this, lol.

All in all I’m liking the improvement :slight_smile:

#5

I think someone else asked a similar question about what you describe in chat; I’ll try to phrase my answer the same way that I did there. Yes, you are correct that real consensus still doesn’t exist for the cube till weight 3, but right now when the score is calculated for player 2, it’s doing a kind of raw comparison and erring on the side of not counting whatever player 2 added. We want to change it to err on the side of counting whatever player 2 added; it just also so happens that it is relatively likely whatever player 2 added will go to the consensus, hence why player 2 will now be awarded about as fairly as player 3. That’s the more precise way of explaining it, but the practical effect is that the result of the scoring calculation for player 2 will be similar to what we do/don’t accept of their trace. Is that clearer?

2 Likes
#6

ooohh ok, so in a way it does invalidate 2 v 1 (without actually doing so lol)

But as you said, what happens if 2nd goes in and fills the cube, isnt that insta 2ks? even if their accuracy will plummet?

#7

The thing about using that cheating mechanism is they would always have to know they were player 2. It’s generally not easy to guess that. Also, if people’s accuracy plummets to a certain extent, the anti-cheating mechanism from 2015 kicks in where they get progressively less and less points. Once we roll this out, we’ll obviously try to keep an eye on people overtracing as a gamble for 2k, but the trick will likely have diminishing returns and in the meantime it seems like the more serious problem is TBers doing skimpier traces that penalize other players points-wise.

#8

true, the only ppl who can know what wt is each cube are scouts + lol and I don’t think anyone scout + would cheat that way.

#9

Yep! And at that point it would also be easier to figure out who was doing that, too.

1 Like
#10

an option would be to inform second player “you are second” (like information you are TBer) and give the points to second after third player has played ?

2 Likes
#11

Like the idea, many games act like that for bonuses and some laters awards :slight_smile:
Perhaps you could also (in addition to @susi idea) adding a “user win x points for 2nd playing” in chat

#12

Think the problem with having to much retros would make people lose a bit of motivation if they get 25 pts every cube they play instead of the real points, if you get them as retros you have no idea of how you got them

1 Like
#13

I’m late to the party here, but I am confused. You say 2nd will be awarded as fairly as 3rd. But the 3rd tracer suffers just as much as the 2nd if there is a bad tb’er. In actually the 3rd tracer gets caught with lower points more, because if they are the tiebreaker between 1st and 2nd, they get the lower # of points. Then, if they agreed with the tb’er the tb’er gets the retros (so the third tracer sees what they should have gotten) if they agreed with the 2nd tracer no one gets anything. So I am confused on how this is supposed to work.

Also, is it in effect currently?

1 Like
#14

It’s not currently in effect, we’re still working on it. As for the bulk of your question-- I talked with Chris, and we’re not sure what you’re describing here in terms of how it causes a problem for the scoring change. Sometimes in the current system Player 3 can be penalized alongside Player 2 by a bad TB, yes. But it can happen that Player 2 adds something to the consensus that the TBer missed, and Player 3 does the same, but only Player 2 is really penalized because Player 2’s score has been calculated against the TB, not against the revised consensus (whereas Player 3 is submitting something that matches the consensus).

Contrast this with if we change scoring to account for things that Player 2 adds, at the time of submission. Then Player 2 should no longer be penalized by a bad TB; the most that could happen is that Player 2 might be unfairly rewarded for adding merger segments after a good TB that didn’t add them. And like I said to Nik, you’d have to know you were Player 2 to intentionally cheat that way, plus your accuracy would suffer in the long run. As for where Player 3 fits into this picture, these revisions will help their points too in tiebreaker scenarios, as long as they’re adding stuff that needs to be added.

1 Like
#15

like it is now, when third player trace correct but either first or second player have done it wrong, third player is still pentalized for that, because the points get awarded before consensus is updated.

2 Likes
#16

Yes, that’s a systemic flaw with the points system. We could only truly solve that by giving all traces retro points and only a very small initial award. Edit: Sorry, reread and I see what you’re saying. Chris’ update below is about that.

One update from Chris, though: it looks like we may have spoken prematurely about making the improvements to benefit Player 2. To quote him:

I was wrong about how the consensus change would affect the second player. The second player behavior will actually remain the same (anything they add that the trailblazer didn’t will be considered overcoloring).

The third tracer behavior does change in that before they were were required to agree with A and B, but now they just have to agree with one of them (majority wins).

We are still testing this, but we hope it will be a welcome change in its own right. As for making changes that can benefit Player 2 against a bad TB, we’ll probably wind up holding a consensus threshold discussion at HQ next week or after the holidays.

Sorry for any confusion, guys!

1 Like
#17

ok now that makes more sense to me lol, ty!

#18

I think a way it could work, would be:

tber submits
2nd player submits doesn’t get pts but gets an estimate based on what’s in consensus from tber, like “this is aprox. ### pts”
3rd player submits, when spawner updates wt to 3 it gives retros to tber, and in a way retros to 2nd and pts to 3rd.

or alternatively, same thing but later on when admin completes cell/cube(s) pts are given based on what’s in consensus in cubes then, if admin has completed then consensus is most likely correct vs at wt 3 where things may still be wrong.

this does have the obvious tradeoff that no insta pts but it does trade that off with more/most accurate pts/retros for everyone.

But it is only 1 idea :slight_smile:

#19

This makes sooooo much more sense!!!

Thank you so much!

1 Like
Update on scoring, accuracy, and Review Mode/Activity Tracker situation
#20

yeah, agree it makes a lot more sense

1 Like