Search the Community
Showing results for tags 'pr'.
Found 3 results
I am curious if there is any interest among the WNx veterans about developing a Player Rating for World of Warships. The two main statistical websites have basically defaulted to using the simplistic %damage, %kills, %win rate as the three data sets they extract from the API. Warships Today rating system (WTR): https://na.warships.today/help/warships_today_rating Wows Numbers rating (PR): https://na.wows-numbers.com/personal/rating Recent work on the wows-numbers site shows their is more data available from the WG api than before - spotting damage, ship spots, kill/death ratio, survival rate, are all showing up in the more comprehensive wows-numbers site, but are not being used in the rating for being superfluous according to the site owner. To a certain extent, I do agree, the best players are the best at PR and WTR based on just those 3 numbers, and while it is abusable by seal clubbing or overplaying ships with low expected values, it's easily detectable by a clan commander. I am more interested in seeing if there is any interest in driving Warships player traffic to this site by developing the preeminent rating system. Maybe everyone is content to let the site fade out as World of Tanks slowly fades from the scene, but there is a very active and dedicated player group for warships, and it has not peeked yet, developing the best rating system would drive traffic here, if that is even something that Solono wants. In any case, I wanted to put this thought out there. I am a mathematical neophyte to be sure, but it is interesting to me to watch how the rating mechanisms come into existence, and it seems like this might be an opportunity to bring new enthusiasm back to wotlabs.
All the figures here will refer to EU clans and ranks. CR = Clan Rating (overall how a clan is doing) wGM = Weighted average of clan's Elo on the global map (how effective the clan is at CW on the global map) I was interested in a clans rank in the league table provided by WG, and how each clan performed on the global map, and the interation between those. I am using ARKT as a baseline, as that is the clan I am presently in, and so watch the most. So last night, 15th September 2015, ARKT lost two CW battles, and lost one province, yet looking at the WG clan league table ARKT went up 18 places from 448 to 430, and CR went up from 8,876 to 8,922 up 46 points. Which is a big jump. Normally I expect a successful landing to give 8-11 points, and a successful defence to give 4-8 points. Average PR only went up one from 6,885 to 6,886. So the increase appears to have nothing to do with what individual members have done by merely playing better. But there was also a big jump in wGM, the factor WG allocates to each clan based on their battlefield success rate on the global map. ARKT went up from 1076 to 1122, which was also a big jump. Last nights ARKT battles have been; 1. Attacked Järpen and lost to [1STLT]at 19.13. 2. ARKT owned province Laxsjö was attacked by [-WIN-] . [-WIN-] won and ARKT lost province Laxsjö. At 19.47. So ARKT fought two battles, one an attack and one a defence, lost both and yet wGM went up 46 points in wGM. Why? At first I just could not understand this. Then I looked at WIN and 1STLT. Both those clans are much higher in the league table than ARKT. 1STLT lost 31 places last night and have dropped down to position 49, while WIN lost 10 places and dropped to position 169. So the explanation is that the formula WG uses has a step in it which effectively rewards clans for fighting clans considerably better than they are, and a penalty for clans that attack clans ranked considerably below them. I had assumed the weaker clan had to actually win to really feel the inequality points bonus, but from last nights result that appears to be not the case. The idea behind that was to stop the best clans repeatedly farming weaker clans. ARKT lost battles and went up 18 places, WIN and 1STLT both won and yet dropped places due to fighting ARKT. But bizarrely if you are a weaker clan, and you went and found 99P on the tier X map (they are presently ranked highest for wGM with a score of 2019) and attacked them the clan would probably get a huge points reward to it's wGM. Is anyone able to confirm or deny this working theory? This lead me off on a tangent to considering the most successful clans on the tier VIII map. Tier VIII map is split into four time zones 5pm, 6 pm, 7pm, and 8 pm. 5pm top three 6pm top three 7pm top three 8pm top three 1. 3XR (1386) TESLA (1275) ZWACK (1276) S4TDD (1339) 2. 1-DPA (1194) BABA (1247) C_R (1257) IDEAL (1310) 3. NE-VO (1192) 1PADX (1241) PSQD (1220) WORKS (1293) Slightly odd how those figures look. Numbers 2 and 3 in the 5pm slot are the weakest. Clans at 1,2, and 3 in the 6pm and 7pm slots are pretty closely ranked. The clans in the 8 pm top three all have higher figures than most of the clans in the earlier time slots (one single exception 3XR). I have not yet worked out a hypothesis for these differences, can anyone assist? Another thing I found out is that WG displays on the global map the wGM averages of clans in tiers. It says Tier VI Tier VIII Tier X 1109 1200 1133 I expected tier VI to be lowest, I was not expecting tier X to be lower than tier VIII. Again does anyone have a plausible explanation for this? So now I got to looking at the tier X map; 5pm top three 6pm top three 7pm top three 8pm top three 1. NOS (1226) S-V (1205) 99P (1346) WHYME (1260) 2. JCKAL (1173) R-BIH (1195) F15 (1299) OMNI (1232) 3. STAR5 (1071) FPP (1179) FAME (1299) NOS (1217) 99P are the highest ranking clan at global wars (EU) (2019 points), however they fight zero stronghold attack/defences. Highest ranking clans in the overall league table in positions 1 and 2 are FAME (1949 wGM points) and OMNI (1848 wGM), and they both fight stronghold attack/defences. Two conclusions 1. If two clans have a rank of the same amount in tier VIII and X on the global map, say for instance 1200, there is more "weight" in being in tier X, and so the wGM will be higher for the clan fighting at tier X even if two clans have 1200. As an example FAME have 1299 at tier X and IDEAL have 1310 at tier VIII, and their respective wGM is 1949 for FAME, and 1310 for IDEAL, a 639 point difference between having the same effectiveness against opponents on the tier VIII map compared to the tier X map. Is there a formula to calculate the precise difference between 1200 on one map and 1200 on another? 2. Stronghold attack/defences give alot of CR points. So 99P fight in the same CW region (tier X) as FAME, in the same time slot (7pm), and clearly do better than FAME at Clan Wars on the global map: 99P first and 1346, against FAME third and 1299, but in the league table of best clans; League Position Clan CR wGM Stronghold 1. FAME 15397 1949 2049 2. OMNI 15305 1848 2280 34. 99P 12442 2019 0 Is there a formula to calculate how many points Stronghold attack/defences give to CR? Link to WG clan rating page; League Position
The spiritual successor to the Project Reality mod from Battlefield 2 is coming soon! The very first videos were released about three months ago and one as recently as December. The very same development team from Project Reality has decided to go commercial with their game on a completely new and top of the line engine. Unreal 4 Engine demonstration: They are in the VERY early pre-Alpha stage of development for the game at this point, but some of the effects such as bullet impacts are just purely amazing. (December Video) (November Animations Video) (1st ever Squad video from October) I'm hyped. Project Reality mod was amazing and the new engine will allow them to do far more than the limited BF2 engine. edit: They just started transitioning today from their old website to a new website worthy of a serious gaming company.