# True Contribution for dummies **URL:** \ **Category:** Tournament\ **Created:** [October 11, 2022, 3:08pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744 "2022-10-11T15:08:11Z")\ **Posts on this page:** 9\ **Page:** 3
**Author:** ![anthill](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/anthill/32/3275_2.png) [@anthill](https://forum.numer.ai/u/anthill)\ **Post date:** [October 18, 2022, 5:50pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/42 "2022-10-18T17:50:08Z")
> [@autratec](#): > > I don’t quite understand the logic and removing and add in that signal will help dramatically improve the portfolio performance, considering CORR/IC has such low score. To give a heuristic example of how you could see a situation where a model gets very low Corr but very high TC, imagine that the market is divided into a number of different sectors and the metamodel does extremely well on all but one of these sectors. If you provide a model which makes very good predictions on that one missing sector, but random predictions on all the other sectors, you will end up with a very low Corr, but your model will nevertheless dramatically improve the performance of the metamodel and will get a high TC. ---
**Author:** ![wigglemuse](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/wigglemuse/32/3582_2.png) [@wigglemuse](https://forum.numer.ai/u/wigglemuse)\ **Post date:** [October 18, 2022, 5:57pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/43 "2022-10-18T17:57:57Z")
Yes, some models can be considered like vitamins for the metamodel – they are just helping supplement some deficiency. ---
**Author:** ![autratec](https://avatars.discourse-cdn.com/v4/letter/a/6bbea6/32.png) [@autratec](https://forum.numer.ai/u/autratec)\ **Post date:** [October 19, 2022, 2:47am UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/44 "2022-10-19T02:47:10Z")
thanks for the explanation. it helps me thinking special strategy to get higher TC, rather than general CORR. ---
**Author:** ![f58c](https://avatars.discourse-cdn.com/v4/letter/f/779978/32.png) [@f58c](https://forum.numer.ai/u/f58c)\ **Post date:** [September 15, 2023, 8:17pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/45 "2023-09-15T20:17:41Z")
@wigglemuse - I’ve uploaded some models and they are now getting tc scores. from day to day, past rounds tc changes and updates. is tc in flux until rounds are “resolved”? my best models tend to stay relatively positive even with these changes, but I’m using tc scores to judge how much NMR to stake. any insight is greatly appreciated! thx! ---
**Author:** ![wigglemuse](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/wigglemuse/32/3582_2.png) [@wigglemuse](https://forum.numer.ai/u/wigglemuse)\ **Post date:** [September 15, 2023, 8:40pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/46 "2023-09-15T20:40:46Z")
All scores on unresolved rounds are in flux – only the final day (day 20) of the round actually counts. The in-progress scores are “as-if” it ended today and are just something to look at. ---
**Author:** ![f58c](https://avatars.discourse-cdn.com/v4/letter/f/779978/32.png) [@f58c](https://forum.numer.ai/u/f58c)\ **Post date:** [September 15, 2023, 9:45pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/47 "2023-09-15T21:45:19Z")
thx! my earliest models will resolve at the end of Sept. it’ll be interesting to see the results. ---
**Author:** ![numerologist](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/numerologist/32/2824_2.png) [@numerologist](https://forum.numer.ai/u/numerologist)\ **Post date:** [September 16, 2023, 6:49pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/48 "2023-09-16T18:49:57Z")
> is tc in flux until rounds are “resolved”? That’s something I don’t understand myself too. One would expect that values would stabilize closer to the end of a round: the farther the round from resolution, the more volatile its TC value (similar behavior we have with CORR). But I checked one of my models which is basically 1-p example predictions, and on Thursday night it had (expectedly) horrible cumulative CORR but a very attractive (literally all the way up) cumulative TC. However, when I checked it on Friday morning, the result for CORR was about the same but all TC values for **all** the unresolved rounds literally **flipped** , i.e. it became cumulative negative TC all the way down. How is something like this even possible **overnight**? It feels like something is really off (a bug?) about the algorithm that computes TC. ---
**Author:** ![gammarat](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/gammarat/32/3281_2.png) [@gammarat](https://forum.numer.ai/u/gammarat)\ **Post date:** [September 21, 2023, 9:11am UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/49 "2023-09-21T09:11:58Z")
> [@wigglemuse](#): > > Leaderboard position is score based only – it used to be corr, now it is TC. (20 round moving average with the current live rounds gaining in weight each day while the 4 least recent of 20 lose weight each day – at least that’s how it was done on corr). The leader board is now based on a 1 year average, and it can be ranked on any of the headings. It just comes up first on TC. ---
**Author:** ![rustydata](https://avatars.discourse-cdn.com/v4/letter/r/9de053/32.png) [@rustydata](https://forum.numer.ai/u/rustydata)\ **Post date:** [August 25, 2024, 12:37pm UTC](https://forum.numer.ai/t/true-contribution-for-dummies/5744/52 "2024-08-25T12:37:15Z")
Wont pretend to fully understand even the intended algo, but I think sharing my understanding of the complexity of tying fund performance may help others. Consider 3 base cases: Most models agree on a prediction and mine does too; Models are fairly distributed on a prediction, mine fell into one decent sized bucket; Most models agree on a prediction, mine does not; Then consider 2 versions of being wrong across those 3 base cases: The fund went with my prediction, but it turned out to be sorta wrong, but only after 15 days, and they arent down, but arent up as much as I predicted. My prediction wasnt wrong, but the existing fund position really didnt allow for it to realize any positive outcome. What complicates this: We are predicting market direction but are being rewarded for fund performance which holds a bunch of proxies and hurdles, not least of which being the difference between market close (train) and what we model (live) vs the fund intraday portfolio and how it actions our models. We arent telling the fund how to enter positions and arent providing confidence intervals for price ranges. We are just giving predictions on a vector (go up a lot, maybe go down), **not** targets. By choosing to stake we are saying we are ok with the layer between. It’s a choice. We can give suggestions, ask questions, scrutinize, and can stake models or start our own funds. [Previous page](https://forum.numer.ai/t/true-contribution-for-dummies/5744.md?page=2)