More tings for Rayman 3 scores!

[Hunchman801]

I disagree, 1000 posts is a lot more effort than a simple 500,000 points. If we are to increase rewards over 500,000, we will decrease those under this threshold.

[rolesfamily]

I disagree, 1000 posts is a lot more effort than a simple 500,000 points. If we are to increase rewards over 500,000, we will decrease those under this threshold.

Then that'd be ridiculous to REDUCE the current payout for people under 500k. At least keep it the same if you have to, don't reduce it.

[Xenon]

Well I'm very happy with the current system, but if it were to be modified, I'd support greater rewards for the upper regions of the HOF, starting at 600 or 700k, not 500k.

[MrBadGuy]

I think we should just get more tings for posts that are awesome.

[rolesfamily]

I think we should just get more tings for posts that are awesome.

Well it looks like I'll have the top rank in no time then!

[Reese Riverson]

I think the current ting earnings per-post are quite sufficient as-is. Awesome post or not. ( Which I don't know why Roles assumes all his posts are all awesome, anyway. )

[rolesfamily]

I think the current ting earnings per-post are quite sufficient as-is. Awesome post or not. ( Which I don't know why Roles assumes all his posts are all awesome, anyway. )

Well that's why I stored my tings temporarily to work through the lower ranks again. I did it twice. It was not as bad as I thought. - And hey you damn well know all my posts are awesome

[Maz]

I recently discovered this topic, and it peaked my interest, as I too think that the amount of Tings given for really good Scores is relatively low. I've also read all the posts here, so let me share my opinion:

First of all, I'm against selecting a certain amount of Points (e.g. the 500.000 which Hunch mentioned), and just saying that everyone below sould get less Tings in relation to his Score than everyone above. Surely you will agree that getting 499.999 isn't all that different from 500.001, so it would be weird to give much more tings to the player who got the latter.

Secondly, rewarding Top-Scores with twice or thrice as much Tings was mentioned. I'm against that as well for an obvious reason; let's say there was a "Player A" who had a Score of 800.000 Points, and a "Player B" whose Score was 850.000 Points. Now A improves to 810.000, and B improves to 851.000, and I Hope you see what I mean by now. Sure, A has improved 10 times as much as B, but it probably was much harder for B to get these 1.000 Points than it was for A to get an extra 10.000. If you want me to bring up an example, I improved from 778.807 to 826.604 within 2 days; now I'm at 847.425 Points, and 850.000 seems almost unreachable.

The two points mentioned above led me to the following conclusion: it would be interesting to have the amount of Tings increase exponentially. That way, "harder improvements" as the one mentioned above are rewarded properly, and noone is put at a disadvantage by selecting a fixed Score for more Tings. As an example, let me bring up an idea that I came up with:

Master said that thrice the current value of Tings would be acceptable for Scores bigger than 800.000, and Adsolution finds an amount of 30k - 40k Tings to be an appropriate reward for such Scores. At the moment, MandM1981 is placed first in the Hall of Fame with 872.222 Points. I hope you will agree with me that 25.000 Tings are quite reasonable for such an achievement. So, here's the first part of my idea:

[1] The person who ranks first in the Hall of Fame shall be awarded 25.000 Tings.

The second part of my idea relates the other Scores to that (this equation is only an example, but I think it would work out pretty well):

[2] Amount of Tings = 25.000 ^ ( [ P + { T - P } / 2 ] / T )

P - Player's Score
T - Top-Score (MandM's 872.222 at hte moment)

Example: Your Score is 654.321 Points.
T - P = 872.222 - 500.000 = 217.901.
{ T - P } / 2 = 217.901 / 2 = 108.951.
P + { T - P } / 2 = 654.321 + 108.951 = 763.272.
[ P + { T - P } / 2 ] / T = 763.272 / 872.222 = 0.875...
25.000 ^ 0.875... = 7.057 = Amount of Tings.

For those interested, that would translate into the following rewards:

872.222 Points --> 25.000 Tings
850.000 --> 21.974
800.000 --> 16.438
750.000 --> 12.297
700.000 --> 9.199
650.000 --> 6.882
600.000 --> 5.148
550.000 --> 3.851
500.000 --> 2.881
450.000 --> 2.155
400.000 --> 1.612
350.000 --> 1.206
300.000 --> 902
250.000 --> 675
200.000 --> 505
150.000 --> 378
100.000 --> 282
50.000 --> 211
0 --> 158

I believe that all Scores below 200.000 can only be reached if either you really try not to gain too many Points, or you just play through the game without caring about Points at all.

Also, it should be mentioned that the current "Amount of Tings = Score / 100" is fulfilled at a Score of 636.688 Points (6.366,89 Tings). Thus, all players with a Score bigger than that would benefit from that, and players with a Score lower than that wouldn't (but there isn't as much of a difference). I personally think that 650k Points can easily be reached by playing the game only a few times, so it doesn't seem too high to me (feel free to disagree with me though). And on the bright side, it may motivate some people on this forum to improve their Scores. I for one would welcome more activity in the Scores-Topic and the Hall of Fame.

Another idea would be to leave the rewards as they are for Scores up to 636.688, and use this equation for Scores bigger than that. In fact, there's endless possibilities to deal with this.

I hope that someone on this forum is still taking this discussion seriously. If so, feel free to add improvements or leave an opinion to my idea. Thanks for taking your time to read this!

[Xenon]

Seems like a really interesting suggestion, Maz. With your configuration, the rewards can increase continuously and without interval (unlike the current system), which is a huge improvement, and the pay-outs do seem more reasonable, too.

However, there are a couple of problems I can see:

1) What happens when/if another player reaches the top spot? Not only will the player in now 2nd position lose a large sum of Tings, it will also affect the rewarding of every other player in the HOF.

2) It could become very difficult to moderate HOF entries as people improve their scores. It would probably require manual validation actually, because I don't think an automated system of this complexity could even be implemented.

What are your views on Hunch's proposal, out of interest?

If anyone else has any comments or suggestions, feel free to post them!

[MandM81]

The suggestion from Maz is good as it certainly diffentiates between the top scores. Maybe one can settle with one variable P instead of two variables T and P?

The modified function

f(P) = 30000^([P+{900000 - P}/2]/900000)

will provide 25k+ Tings for the current top score and it eliminates the need for adjusting the number of Tings for everyone else when a new best score has been reached.

The top scorer doesn’t necessarily get a nice round number of Tings, but one can probably live with that.

Also, I assume the score 900k will not be bettered. I may be wrong about that, but one can adjust the function accordingly. A score of 900k points will lead to 30.000 Tings.

Now a tricky question, should the number Tings be adjusted for the use of IPG?

[saerleiya]

It depends if the IPG has a big impact or not on the scores, for example if you can reach very unlikely scores in normal conditions with it. Otherwise, how can you really proove if it was used or not to obtain 750k?

[Maz]

One might of course do it differently, for example, set the reward to 15.000 - 20.000 Tings for 800.000 Points and use that as a starting point instead of the Top-Score, e.g.:

Amount of Tings = 17.000 ^ ( [P + { 800.000 - P } / 2 ] / 800.000 )

The idea behind using the Top-Player as a standard was the following: If MandM improved his Score to, say 873.000, then everyone with a Score less than that would get further away from what we know to be the highest reachable Score, thus, 500.000 Points for example would be worth slightly less. In hindsight, it's a stupid idea though.

The problem with having the amount of Tings increase exponentially is that low Scores also mean pretty low rewards (otherwise, either the equation would become to complex, or there would be ridiculously high rewards for good Scores). To answer your question, for that reason, I really like Hunch's Suggestion, especially in combination with this idea; one could use the old way up to a certain amount of Points, and then start using the equation above once that Score is reached. In the example above, that Score should be 639.400 Points, since the current reward and the reward using the equation would be the same, namely 6.394 Tings, which allows for a "silent transition".

I have to admit that I'm not an expert when it comes to programming, but what about this idea (I'll use the example above; also, let's assume that there's 2 players - A and B - who have Scores of 500.000 and 700.000 respectively):

First of all, a variable could be created for an account which saves the amount of Tings gotten from the HoF. As soon as you enter your code, the following things might happen:

• Set that variable to Zero.
• Something like "If Score < 639.400 then Tings_HoF := Score div 10", which is true for player A, so he'll get 500.000 / 100 = 5.000 Tings. However, it's not true for player B.
• "Else Tings_HoF := equation above;", which IS true for player B, so he'd get 9.247 Tings.

Now if B improved his Score to 800.000, step one would set Tings_HoF back to Zero before calculating the new value. As I said, I'm not an expert on programming, but would it really be that hard to implement?

Edit, because M&M replied before I could finish this post:

That's a tricky question, but since the IPG has been allowed into the HoF, it is a tool equal to anything else, and should be unimportant for the reward. People like you and me will most likely never be friends with the IPG, but I guess that's our problem then, since we're not exaxctly forced into not using the IPG. Well, for the time being, it's not that important either way, since DTUCC seems to be the only player who would benefit from it at the moment. But that's just my two cents, feel free to share your opinion on that matter.

[MandM81]

It depends if the IPG has a big impact or not on the scores, for example if you can reach very unlikely scores in normal conditions with it. Otherwise, how can you really proove if it was used or not to obtain 750k?

That's a tricky question, but since the IPG has been allowed into the HoF, it is a tool equal to anything else, and should be unimportant for the reward. People like you and me will most likely never be friends with the IPG, but I guess that's our problem then, since we're not exaxctly forced into not using the IPG. Well, for the time being, it's not that important either way, since DTUCC seems to be the only player who would benefit from it at the moment. But that's just my two cents, feel free to share your opinion on that matter.

I was actually kidding about the IPG adjustment. Just seeing the formula put me in math mode. It could a fun challenge from math point of view to concoct a formula that took into account the different strategies and platforms.

[Hunchman801]

First of all, I'm against selecting a certain amount of Points (e.g. the 500.000 which Hunch mentioned), and just saying that everyone below sould get less Tings in relation to his Score than everyone above. Surely you will agree that getting 499.999 isn't all that different from 500.001, so it would be weird to give much more tings to the player who got the latter.

I don't think you understood my suggestion, which was not to create a gap in the rewards, but to modify to formula to give less Tings than the previous one for scores under 500,000 and more than before for the others. The function would remain continuous.

With your configuration, the rewards can increase continuously and without interval (unlike the current system), which is a huge improvement, and the pay-outs do seem more reasonable, too.

The currently system is actually based on a linear reward of 0.01 Ting per point, which is continuous and without interval (let alone the final rounding to get a even Ting amount).

Another idea would be to leave the rewards as they are for Scores up to 636.688, and use this equation for Scores bigger than that.

As explained in my suggestion, I am strongly opposed to such an idea.

To answer your question, for that reason, I really like Hunch's Suggestion, especially in combination with this idea; one could use the old way up to a certain amount of Points, and then start using the equation above once that Score is reached.

I have no idea what you're referring to as my suggestion but it was the exact opposite of this idea.

As for a better formula, I think Maz came up with a nice suggestion, but it should not be based on the top score for two reasons. First, this means we won't have to recalculate (and lower!) everyone's reward when a new top score is reached, and second, we won't have to store the Ting reward anywhere as the old and new scores will be enough to calculate the difference between the old and new rewards.

When it comes to the formula itself, I'd suggest a more scientific approach. First come up with a model, and then, calibrate it. The current model is A . eP/B, so all we need is two pairs of score and reward and we'll get the constants for the formula.

We could also use a polynomial model instead of an exponential one to minimise the impact on low scores, even a simple one like A . PB (this one in particular would also solve the issue where a score of zero gives a non-zero reward).

[Maz]

I just noticed that I completely misread your post... my apologies!

As I said, the formula was just an example. My point was that I'd like a formula like x^y, partly for better rewards when improving Scores in higher regions. I have nothing against a polynomial model either, but I'd just like to point out that a Score of 0 Points can't be achieved anyways. I still believe that a Score of less than ~150.000 - 200.000 Points can only be obtained if you don't care about the Points at all, or try to avoid getting Points, but whatever.

That still leaves the question on what the actual formula should look like.

[Xenon]

With your configuration, the rewards can increase continuously and without interval (unlike the current system), which is a huge improvement, and the pay-outs do seem more reasonable, too.

The currently system is actually based on a linear reward of 0.01 Ting per point, which is continuous and without interval (let alone the final rounding to get a even Ting amount).

Eek, didn't even realise you have since implemented that system. I thought we were still using my setup. I guess that shows how out of touch I now am with this place!

[PluMGMK]

What system was that? I seem to have forgotten!

[Xenon]

Here is the former system. Hunch transformed it into an automated system paying members 0.01 Tings per point, as mentioned in this thread, in late 2012 it seems. Christ