Search the Community
Showing results for tags 'data'.
Found 4 results
Hey guys, I dont know if this is the right place to post this, but i am trying to build a spreadsheet so it helps my clan with some basic things The idea is you input the name of the player, and it automatically loads you the basic stats of said player. What i've could do until now is this https://docs.google.com/spreadsheets/d/1fo5JsNNTRpH0mN1xHTN58CV9u4ECCEX4c_rqF-aG8A8/edit?usp=sharing The problem is that i used a function called "IMPORTXML" to retrieve the data from Wotlabs but it seems that the google spreadsheets have a limitation on how many of this formulas you use. With a few players this was working like a charm, but once i started adding up more players, the function keeps saying "loading" and it takes a looot of time to fill the cells. Any idea or help with how i can fix or do this? Thanks guys!
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.
I stopped playing tournaments when WG nerfed the gold payouts. By March of 2016, WG changed how tournament data was presented and added an API. That broke the data scraping scripts that I had running, and I didn't care to update them. Now a year and a half later, I'm taking down the server that hosted the website. I really liked how people used WoTScout.com links in forums to show their past performance. That was a neat side-effect of what was initially just a way for me to see the WN8 and win rate of upcoming opponents during tournaments. Since that historical data may be of value to someone, I'm publicly sharing a full dump of the PostgreSQL database that backed WoTScout.com. Please feel free to do what you will with it. wotscout.sql.tar.bz2 To make the data more accessible, I'll also publishing the Python/SQLAlchemy models (complete with the hard-coded images that I added to @Mort's and @CarryMode's team pages ages ago). models.py
From: http://forum.worldofwarships.eu/index.php?/topic/59528-armor-penetration-curves/ fnord_disc over at EU managed to cobble together code that fits very closely to what WG uses to calculate penetration over distance, and recently released the source code for it (which @Captain_F22ACE improved upon, more on that below). I won't get into the specifics of the equation (feel free to read the reddit link below for that), and a lot of this data requires the ability to extract gameparams.data to obtain specific shell characteristics not listed in-game such as drag constant and krupp value. Basically how to read the graph: Pensacola/Indianapolis/New Orleans 203mm AP: Blue line = raw penetration (what we are familiar with in WoT since shell arc is minor outside of derps and SPGs). Raw penetration assumes you hit a perfectly flat, 0 degree surface Red line = penetration against the belt armor of the ship with vertical impact angle adjusted (but horizontally, e.x. player-controlled angling by turning the ship, still being 0) The creator published this code on reddit: https://www.reddit.com/r/WorldOfWarships/comments/560yg2/wows_ballistic_model_penetration Fun fact from reddit: https://www.reddit.com/r/WorldOfWarships/comments/560yg2/wows_ballistic_model_penetration/d8flmgs Also, @Captain_F22ACE improved upon this code. Added simulated penetration after striking an armor plate. I'll let him post the source code himself in case he doesn't want others to see But it looks something like this (Iowa 16" SHS AP shell used for example, sorry for overlapping text because 14" laptop scaling on 1080P does screwy things) ^155mm thick plate used here btw (Mosvka belt armor), and also with scout plane + main battery +16% range module included so you can see where plunging fire starts to take effect. Gun elevation sadly never reaches full 45 degrees like it historically did, maxes out to 30 instead. Distance is read in meters x 10^4, so "1.5" on the x-axis = 15km, "2" = 20km, and so on. ... It's not perfect since shells of very low mass (76mm on Smith) tend to get quirky readings on the improved code for calculating energy left after hitting the first place (as ships do have multi-layered armor, so citadeling sometimes requires going through 3-4 plates of armor), but the raw penetration calculations are still relatively accurate: ^ 30mm plate used here btw, didn't set actual gun angle for this so range is iffy. So yeah, have fun pouring over the graphs already there and seeing what is the optimal distance to engage a target.