| Author | Thread |
|
|
04/10/2013 04:40:10 PM · #201 |
|
|
|
04/10/2013 04:41:28 PM · #202 |
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
|
|
|
04/10/2013 04:45:54 PM · #203 |
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
|
|
|
04/10/2013 04:46:02 PM · #204 |
Originally posted by mike_311:
do you think its easier to get 2nd place out of 50 or 8th place out of 200? both hold the same placement. |
They are the same, both placements are of equal difficulty.
Actually I take that back, 3rd out of 50 and 8th out of 200 are of equal difficulty, both are 96%. Your example came out to 98% vs 96%
Message edited by author 2013-04-10 16:52:39.
|
|
|
|
04/10/2013 05:21:48 PM · #205 |
Originally posted by MAK:
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
I hope next season it remains random! The whole idea of stacked teams irks me.
The system in place with scoring this year works well......I don't feel it needs changing. In the beginning the members running this TPL stated that the main outcome was to increase participation and it is.
Your team should encourage each other to enter more challenges. This is the best way to enjoy the whole competition.
|
|
|
|
04/10/2013 05:28:18 PM · #206 |
Originally posted by MAK:
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
And I like very much that they were, on the whole, randomly generated. I am thoroughly enjoying getting to know some very nice folks I would probably never have met had we not been randomly placed on the same team. If we do it again, I would like to get to know more and different people - not because I haven't had fun and wouldn't like to work with these people again, but simply in order to broaden my base of DPC friends. Side challenges can do that, too. |
|
|
|
04/10/2013 06:13:32 PM · #207 |
Originally posted by bhuge:
Originally posted by mike_311:
do you think its easier to get 2nd place out of 50 or 8th place out of 200? both hold the same placement. |
They are the same, both placements are of equal difficulty.
Actually I take that back, 3rd out of 50 and 8th out of 200 are of equal difficulty, both are 96%. Your example came out to 98% vs 96% |
You must be using "new" math...
Anyway, statistically both are equal. But in real life there if there 200enttries. 150 will not be of a high placement quality. Please dont anyone take that the wrong way. its just the way it is, you need to buffer for mediocity.
When I run my golf league we don't set handicaps to ensure equality. We set them to ensure competition. If I shoot a 94 avg and you shoot a 80. Your handicap would be 7 and mine would be 19. Not 8 and 22. There is a factor to ensure it isn't unfair to the better player. You make it so the lesser golfer can compete and strive to improve while giving the better golfer his natural advantage and not put him at a disadvantage just because you can. |
|
|
|
04/10/2013 06:14:36 PM · #208 |
Originally posted by nam:
Originally posted by MAK:
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
And I like very much that they were, on the whole, randomly generated. I am thoroughly enjoying getting to know some very nice folks I would probably never have met had we not been randomly placed on the same team. If we do it again, I would like to get to know more and different people - not because I haven't had fun and wouldn't like to work with these people again, but simply in order to broaden my base of DPC friends. Side challenges can do that, too. |
I agree. That's probably the best part of this whole experience. The interaction is great. |
|
|
|
04/10/2013 06:33:56 PM · #209 |
Originally posted by mike_311:
Originally posted by nam:
Originally posted by MAK:
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
And I like very much that they were, on the whole, randomly generated. I am thoroughly enjoying getting to know some very nice folks I would probably never have met had we not been randomly placed on the same team. If we do it again, I would like to get to know more and different people - not because I haven't had fun and wouldn't like to work with these people again, but simply in order to broaden my base of DPC friends. Side challenges can do that, too. |
I agree. That's probably the best part of this whole experience. The interaction is great. |
I agree also; it makes for a better experience all around. The way Cory's comment was worded seemed to imply that we should "choose" teams - not what I think should be done. It is great meeting new folks and the different styles they bring to the table.
|
|
|
|
04/10/2013 06:45:38 PM · #210 |
Originally posted by mike_311:
Originally posted by bhuge:
Originally posted by mike_311:
do you think its easier to get 2nd place out of 50 or 8th place out of 200? both hold the same placement. |
They are the same, both placements are of equal difficulty.
Actually I take that back, 3rd out of 50 and 8th out of 200 are of equal difficulty, both are 96%. Your example came out to 98% vs 96% |
You must be using "new" math...
Anyway, statistically both are equal. But in real life there if there 200enttries. 150 will not be of a high placement quality. Please dont anyone take that the wrong way. its just the way it is, you need to buffer for mediocity.
When I run my golf league we don't set handicaps to ensure equality. We set them to ensure competition. If I shoot a 94 avg and you shoot a 80. Your handicap would be 7 and mine would be 19. Not 8 and 22. There is a factor to ensure it isn't unfair to the better player. You make it so the lesser golfer can compete and strive to improve while giving the better golfer his natural advantage and not put him at a disadvantage just because you can. |
I'm not sure how you calculate percentages but DPC appears to do it the same way I do....
Needle Pulling Thread Macro II 09/21/2003 6.975 8 / 200 96% 33
A Girl and Her Bunny Kids With Toys 12/26/2010 6.9935 3 / 50 96% 37
dolly done Kids With Toys 12/26/2010 7.2403 2 / 50 98% 61
I don't understand your "buffer for mediocity". Are you saying when you increase the number of entries you only add crappy photographers?
Golf handicaps are also a whole other story, they also don't change based off the number of players (I think?, not a golfer) so I don't see how they relate to this topic.
ETA: fix typo
Message edited by author 2013-04-10 18:46:34.
|
|
|
|
04/10/2013 07:25:41 PM · #211 |
well thats interesting 2/50 is 4 out of 100 or 4% or 96% percentile... oh well, no big deal.
no im not saying that that adding more only adds crappy photographers, but lets face it, unless your image stands out enough to get a higher score its going to get lost in that sea of mediocrity. again im not trying to make any point about quality or anything only that its easier to rise above a whole bunch of ok shots than it is to rise above the top of a lesser amount of shots.
the point with the handicaps is that same thing, the images that place lower should not carry the same weight as ones who place higher.
tacking My personal example i gave earlier, my 7th place entry is far worse statistically than my 19th place entry, how its that fair? even though it placed higher and score much higher. because of the number of entries its not worth as much, not even by a close amount...
|
|
|
|
04/10/2013 07:43:32 PM · #212 |
Originally posted by grahamgator:
Originally posted by mike_311:
Originally posted by nam:
Originally posted by MAK:
Originally posted by grahamgator:
Originally posted by Cory:
I'd also like to point out that during the next season choosing your teammates wisely is going to be much more important, even without a change to the scoring system. ;) |
I thought teams were "randomly" chosen this time.?? |
Correct. |
And I like very much that they were, on the whole, randomly generated. I am thoroughly enjoying getting to know some very nice folks I would probably never have met had we not been randomly placed on the same team. If we do it again, I would like to get to know more and different people - not because I haven't had fun and wouldn't like to work with these people again, but simply in order to broaden my base of DPC friends. Side challenges can do that, too. |
I agree. That's probably the best part of this whole experience. The interaction is great. |
I agree also; it makes for a better experience all around. The way Cory's comment was worded seemed to imply that we should "choose" teams - not what I think should be done. It is great meeting new folks and the different styles they bring to the table. |
Interesting. Love to hear that you guys liked the random teams. Perhaps we won't implement the change that we discussed if the demand is this strong to remain random. Much is still up in the air - thanks for all the feedback everyone! |
|
|
|
04/10/2013 07:43:59 PM · #213 |
Originally posted by mike_311:
well thats interesting 2/50 is 4 out of 100 or 4% or 96% percentile... oh well, no big deal.
no im not saying that that adding more only adds crappy photographers, but lets face it, unless your image stands out enough to get a higher score its going to get lost in that sea of mediocrity. again im not trying to make any point about quality or anything only that its easier to rise above a whole bunch of ok shots than it is to rise above the top of a lesser amount of shots.
the point with the handicaps is that same thing, the images that place lower should not carry the same weight as ones who place higher.
tacking My personal example i gave earlier, my 7th place entry is far worse statistically than my 19th place entry, how its that fair? even though it placed higher and score much higher. because of the number of entries its not worth as much, not even by a close amount... |
Percentages get weird when applied to placement. You need to shift them by one to get the right number. Think about the extreme case of a challenge with only one entry. Does it get 100% or 0%. Statistically speaking it gets 50% because it's the only entry, but if we calculated it that way, even first place wouldn't be 100%. It's weird... but probably irrelevant to the actual discussion :)
Sorry I didn't catch the start of the conversation, but now that I see your two entries I think it might be somewhat of a corner case that's hard to capture with any scoring algorithm. The comparison is special because of the Nude challenge. It is unique because it cuts out a lot of the hobby photographers that historically score lower. The majority of entries ended up being from professionals that have access to models that are willing to do a nude photo. So I don't think the percentage discrepancy is due to the number of entries but instead is due to the unique resources that the entrants had.
So while I agree that the percentage score may not have been fair between these two challenges, I don't think weighting the numbers based off participation is the right way to address it.
|
|
|
|
04/10/2013 08:34:28 PM · #214 |
Originally posted by Epsi:
Originally posted by mike_311:
the more i look at week two, the more i think you guys need to update the scoring system next time, it is completely wrong we got zero points for our phenomenal showing.
maybe award the points for individual matches (3, 1.5, 0) and then bonus points for a team win? |
If you mean using the sum of the individual scores as the final match score, I would be keen on removing the limitation of the four top scores and consider all 6 individual matches, since an individual match would have little weight in the global result and would allow all the members of the team to contribute equally to the score. Also the bonus would have to be increased.
In the particular example that you mention it would be like this:
LydiaToo vs. jagar
Things In A Row:97.66% = 1.5pts Things In A Row:94.53% = 1.5pts
mike_311 vs. PennyStreet
Make Believe Sports:97.06% = 1.5pts Things In A Row:92.97% = 1.5pts
CNovack vs. colorcarnival
Lucky 13 II:85.94% = 1.5pts Lucky 13 II:84.38% = 1.5pts
mariuca vs. Yandrosxx
Things In A Row:70.31% = 0pts Lucky 13 II:80.47% = 3pts
hajeka vs. Paul
Lucky 13 II:56.25% = 0pts Make Believe Sports:79.41% = 3pts
RyanWareham vs jaysonmc
Things In A Row:28.91% = 0pts Things In A Row:62.5% = 3pts
Final score: 4.5 + Bonus 13.5 + Bonus
|
If I understand correctly, the 5% adjustment rule is to balance out the uneven number of entries in the different challenges.
So... why does it also apply to the event when both contenders have their high scores in THE SAME CHALLENGE? I'm sure it would be hard to write the code for it, but... if one person has a higher percentage finish in the SAME challenge as their competition, it seems that it's fair to take those percentages as final without the 5% buffer. After all, since they're in the same challenge, all things are equal and there doesn't need to be something like a 5% difference bonsu to make up the difference in the number of entries when... there is no difference.
Right?
|
|
|
|
04/10/2013 08:47:10 PM · #215 |
anther great point Lydia, one I overlooked previously. a handicap was applied when it should not have been. we had two straight out wins that we did not get because of the 5% rule being applied when it no reason to be.
if anything this change out to be made. |
|
|
|
04/10/2013 09:00:13 PM · #216 |
Another vote for random team assignments. More people getting to know each other, helping each other out, cheering each other on ... it really strengthens the community aspect of DPC. IMO.
Also, since my team is leading the Eastern League, the current scoring system must be absolutely brilliant ;-). Although I concede Lydia's suggestion makes a great deal of sense. So maybe some minor tweaks before next season. Without losing the brilliant bits. |
|
|
|
04/10/2013 09:59:21 PM · #217 |
Originally posted by LydiaToo:
So... why does it also apply to the event when both contenders have their high scores in THE SAME CHALLENGE? I'm sure it would be hard to write the code for it, but... if one person has a higher percentage finish in the SAME challenge as their competition, it seems that it's fair to take those percentages as final without the 5% buffer. After all, since they're in the same challenge, all things are equal and there doesn't need to be something like a 5% difference bonsu to make up the difference in the number of entries when... there is no difference.
Right? |
Originally posted by mike_311:
anther great point Lydia, one I overlooked previously. a handicap was applied when it should not have been. we had two straight out wins that we did not get because of the 5% rule being applied when it no reason to be.
if anything this change out to be made. |
so... let me see if I understand your argument. two challenges one 100 entries and other 99 entries and when you compare with in the same challenge, what do you do? no tie? tie consecutive positions (1%)? and when you compare different challenges then you use 5%, despite there is only one entry of difference? Oh maybe you only apply the 5% when there is some number of entries of difference, what is this number? 10, 20 or is also a percentage? and if it is a percentage is it a percentage of the large challenge or the short challenge?
I ask again, try to think on the scoring system before it is proposed. Make logical arguments with actual examples (data of the weeks 1 - 2 and 3 is publicly available for you to use). If what you want is to vent don't expect a reply.
|
|
|
|
04/10/2013 10:11:26 PM · #218 |
referring to Lydia's post, she took 4th and Jagar took 8th IN THE SAME CHALLENGE yet they tied in TPL scoring. In the other, Chrystyne took 19 and Colorcarnival took 21st IN THE SAME CHALLENGE, close, but still she place higher... if these were different challenge you can argue the 5% rule, but in the same challenge?
just pointing out an obvious flaw there...
Originally posted by Epsi:
I ask again, try to think on the scoring system before it is proposed. Make logical arguments with actual examples (data of the weeks 1 - 2 and 3 is publicly available for you to use). If what you want is to vent don't expect a reply. |
no one is venting, just pointing out that he scoring system probably could use a tweak. i accept is is what it is. i haven't posted a concrete solution because i don't have one and I'm hoping by discussing it out loud we can help to find one and if i can come up with a workable method i pull put number is to it.
no one is criticizing your method, it was a very good system but you couldn't possibly have foreseen every possible circumstance where it needs improvement.
Message edited by author 2013-04-10 22:17:47. |
|
|
|
04/10/2013 10:22:49 PM · #219 |
Originally posted by mike_311:
referring to Lydia's post, she took 4th and Jagar took 8th IN THE SAME CHALLENGE yet they tied in TPL scoring. In the other, Chrystyne took 19 and Colorcarnival took 21st IN THE SAME CHALLENGE, close, but still she place higher... if these were different challenge you can argue the 5% rule, but in the same challenge?
just pointing out an obvious flaw there...
Originally posted by Epsi:
I ask again, try to think on the scoring system before it is proposed. Make logical arguments with actual examples (data of the weeks 1 - 2 and 3 is publicly available for you to use). If what you want is to vent don't expect a reply. |
no one is venting, just pointing out that he scoring system probably could use a tweak. i accept is is what it is. i haven't posted a concrete solution because i don't have one and I'm hoping by discussing it out loud we can help to find one and if i can come up with a workable method i pull put number is to it.
no one is criticizing your method, it was a very good system but you couldn't possibly have foreseen every possible circumstance where it needs improvement. |
No, you've nailed what was missing from her post - a concrete example, and proof that it is really a possible issue. (besides, all dictators have cranky moments, last week was mine. ;) )...
We'll definitely add this to the list, we've got a lot to hash out still on this. If you see problems, don't hesitate to mention them, just make sure that you can show concrete proof of the issue.
Now - here's a question to everyone - we need your input about the transition phase. How would you like to see us transition?
All teams go free-agent and can trade as they see fit?
All teams drop half the team into the pool, from which they are reassigned?
All teams drop everyone into the pool, captains remain captains?
All teams dissolve, new teams?
Message edited by author 2013-04-10 22:24:08. |
|
|
|
04/10/2013 10:52:34 PM · #220 |
here is an idea. keep the percentage system, only use it like this:
percent variance = sqrt(difference in entries)/2
this yields:
difference in entry - % variance
0 - 0
5 - 1.1
10 - 1.6
20 - 2.2
30 - 2.7
40 - 3.2
50 - 3.5
60 - 3.9
70 - 4.2
80 - 4.5
90 - 4.7
100 - 5.0
110 - 5.2
120 - 5.5
130 - 5.7
140 - 5.9
150 - 6.1
160 - 6.3
170 - 6.5
180 - 6.7
190 - 6.9
200 - 7.1
so basically the same challenge wont have any possible variance, those with a small disparity or none will get a small variance 1-3% and those with a higher disparity get a greater variance.
in my example earlier where the nude and duotone entry the adjustment would have been 6% it would have been close but the 7th place still would have lost to the 19th place, but at least it would have had a shot and had it placed one spot better it could have over taken it.
Message edited by author 2013-04-10 22:56:42. |
|
|
|
04/10/2013 11:04:47 PM · #221 |
i m kinda enjoying the TPL. i m shooting - under specific conditions, while trying to keep in mind that the crowd likes specific things, and trying to execute well within the schedule. its not something i do often - it is something that can be frustrating, but its churning out some ideas i never knew i had.
edit: meant to add that its quite different than the assignments i usually get - so its educational in a different way. Also liking my teams inputs and conversation.
Message edited by author 2013-04-10 23:10:31. |
|
|
|
04/10/2013 11:06:37 PM · #222 |
| I have really enjoyed getting to know new DPCers, so my vote would be for a complete dissolution and another random selection process. |
|
|
|
04/10/2013 11:11:19 PM · #223 |
Originally posted by mike_311:
here is an idea. keep the percentage system, only use it like this:
percent variance = sqrt(difference in entries)/2
this yields:
difference in entry - % variance
0 - 0
5 - 1.1
10 - 1.6
...
170 - 6.5
180 - 6.7
190 - 6.9
200 - 7.1
so basically the same challenge wont have any possible variance, those with a small disparity or none will get a small variance 1-3% and those with a higher disparity get a greater variance.
in my example earlier where the nude and duotone entry the adjustment would have been 6% it would have been close but the 7th place still would have lost to the 19th place, but at least it would have had a shot and had it placed one spot better it could have over taken it. |
I like this!
There should be NO variance adjustment on entries from the same challenge.
There should be varying degrees of variance in direct proportion to the variance of the amount of entries in the various challenges, so there's not an unfair "give" of the 5% rule to challenges with not much variance in the number of entries.
Providing a 5% variance to the lesser percentage scorer on the same challenge isn't cool. There is nothing to correct in a head-to-head competition.
Keep in mind that I am NOT complaining about the scoring for this TPL. We all knew the rules before we entered.
I am suggesting things for the next round...
I am exceedingly glad that our cohorts have put this TPL together for us and to draw in more entries.
I think it's working.
I know for sure that I LOVE the random draw of the teams to make them more "equal" and that none of us (except the last teams) got to be on a team with anyone that they requested to have on their team. I also like it that (apparently) none of us had to have the ones we requested that we didn't have on our teams. (That would have frustrated people and made for less participation.)
I also love it that the "producers" of our fun time here on TPL are willing to listen to ideas that will make TPL even better,
Life is good.
Let's have... FUN!!!
|
|
|
|
04/10/2013 11:11:23 PM · #224 |
Originally posted by Cory:
Now - here's a question to everyone - we need your input about the transition phase. How would you like to see us transition?
All teams go free-agent and can trade as they see fit?
All teams drop half the team into the pool, from which they are reassigned?
All teams drop everyone into the pool, captains remain captains?
All teams dissolve, new teams? |
I'm a bit hesitant for straight free agent because then we enter into Ribbon Hogs territory, and while I don't think we need to completely flatten the field, per se, I think some constraints for weighting are good and necessary to keep the league worthwhile for the average user.
I'm also a bit against dropping half the team because that will seem a bit crappy for whoever gets dropped and also likely relate to the first issue I noted. I think an all or nothing for team retention is best so far as that is concerned.
As a captain, I'm indifferent to the retention of captains and a selection from there. It did seem to work alright with the DPL, but again, we'd need a weighting system and I've no clue how that all worked. |
|
|
|
04/10/2013 11:17:35 PM · #225 |
Originally posted by mike_311:
here is an idea. keep the percentage system, only use it like this:
percent variance = sqrt(difference in entries)/2
this yields:
difference in entry - % variance
0 - 0
5 - 1.1
10 - 1.6
20 - 2.2
30 - 2.7
40 - 3.2
50 - 3.5
60 - 3.9
70 - 4.2
80 - 4.5
90 - 4.7
100 - 5.0
110 - 5.2
120 - 5.5
130 - 5.7
140 - 5.9
150 - 6.1
160 - 6.3
170 - 6.5
180 - 6.7
190 - 6.9
200 - 7.1
so basically the same challenge wont have any possible variance, those with a small disparity or none will get a small variance 1-3% and those with a higher disparity get a greater variance.
in my example earlier where the nude and duotone entry the adjustment would have been 6% it would have been close but the 7th place still would have lost to the 19th place, but at least it would have had a shot and had it placed one spot better it could have over taken it. |
I don't disagree with your proposal, but just to play Devil's... Isn't it the team's strategic choice to stay out of a challenge with few entries? What I mean is that a team not entering a challenge with few entries is not in their strategic best interest as they strengthen the scoring of others who have entered. Knowing this, why aren't more people entering low # challenges? Hypothetically speaking... consider the difficulty to make  vs 
I'm not saying one photo is better than the other, or even that an entry should be rewarded for it being "hard" to create, but consider that the challenges with few entries have few entries because they are difficult topics or editing sets, they have their own difficulty modifier. Just something to mull over, I guess. |
|
