|
On March 22 2023 19:32 Dakota_Fanning wrote:Show nested quote +On March 22 2023 16:17 MeSaber wrote:Rant: In Players tab, i would like to exclude players based on if they dont play BGH+Low, ie i want to find BGH or Low only players, maybe even be able to set a percent of played map type like 10%. Some have 1% or less of BGH/Low and i dont count those as BGH/Low players. In Games tab, i am trying to find players of 3v3 and 4v4 games AND only BGH/Low maps. Also i dont get any help how to use regexp map. If i could regexp the 'matchup' to include both 3v3 and 4v4 matchups like: XXXvXXX, XXXXvXXXX, ie wildcard a race and amount of players. Or 'format' be able to choose multiple options. Avg APM seems pretty useless, would rather have Avg EAPM. Avg APM can fluctuate a lot per player but EAPM usually does not. I would really like if you could use regexp in every field to find what ure looking for. This would require a regexp help-popup how regexp works ofc as most ppl dont know about it but its a strong tool. So my current confusion is how i find BGH/Low games with 3v3 or 4v4 format included. The lack of sorting is an issue too. It would be very useful. Edit: Be able to exclude short games like less than 5 minutes. Maybe repmastered should remove invalid games that are less than 2 minutes long (any game below 2 min is a disc). https://repmastered.app/?from=2021-01-01&mapMatch=substr&map=hunter&format=4v4&avgAPM=300 as an example. Thx for making this awesome website. Thanks for the feedback. I'd rather answer in general that addresses most of your issues / requests. Please keep in mind that performance is one of the first prio. If you have 10 replays, can you form any kind of query and sortig, the results will come in an instance. repmastered.app has more than 1.6 million replays as of now, and this number just keeps growing. To have "instant" results, the data has to be indexed in multiple ways to support queries and sorting. Building and maintaining indices is a significant task and space, and there are certain limitations. Performance (execution time) of queries and sorting will only be acceptable if they are backed by indices. Even though I would want to allow querying and sorting by all values, this is not possible in practice and is not possible due to database limitations. I carefully select filters and sort orders that seem useful / reasonable. I'll keep revisiting this issue from time-to-time, but I can't just allow all because without index support, filtering and sorting from 1.6 million replays would take too long and would overload the server.
Couldnt the sorting be client-side?
|
Dakota_Fanning
Hungary2345 Posts
On March 23 2023 15:18 MeSaber wrote:Show nested quote +On March 22 2023 19:32 Dakota_Fanning wrote:On March 22 2023 16:17 MeSaber wrote:Rant: In Players tab, i would like to exclude players based on if they dont play BGH+Low, ie i want to find BGH or Low only players, maybe even be able to set a percent of played map type like 10%. Some have 1% or less of BGH/Low and i dont count those as BGH/Low players. In Games tab, i am trying to find players of 3v3 and 4v4 games AND only BGH/Low maps. Also i dont get any help how to use regexp map. If i could regexp the 'matchup' to include both 3v3 and 4v4 matchups like: XXXvXXX, XXXXvXXXX, ie wildcard a race and amount of players. Or 'format' be able to choose multiple options. Avg APM seems pretty useless, would rather have Avg EAPM. Avg APM can fluctuate a lot per player but EAPM usually does not. I would really like if you could use regexp in every field to find what ure looking for. This would require a regexp help-popup how regexp works ofc as most ppl dont know about it but its a strong tool. So my current confusion is how i find BGH/Low games with 3v3 or 4v4 format included. The lack of sorting is an issue too. It would be very useful. Edit: Be able to exclude short games like less than 5 minutes. Maybe repmastered should remove invalid games that are less than 2 minutes long (any game below 2 min is a disc). https://repmastered.app/?from=2021-01-01&mapMatch=substr&map=hunter&format=4v4&avgAPM=300 as an example. Thx for making this awesome website. Thanks for the feedback. I'd rather answer in general that addresses most of your issues / requests. Please keep in mind that performance is one of the first prio. If you have 10 replays, can you form any kind of query and sortig, the results will come in an instance. repmastered.app has more than 1.6 million replays as of now, and this number just keeps growing. To have "instant" results, the data has to be indexed in multiple ways to support queries and sorting. Building and maintaining indices is a significant task and space, and there are certain limitations. Performance (execution time) of queries and sorting will only be acceptable if they are backed by indices. Even though I would want to allow querying and sorting by all values, this is not possible in practice and is not possible due to database limitations. I carefully select filters and sort orders that seem useful / reasonable. I'll keep revisiting this issue from time-to-time, but I can't just allow all because without index support, filtering and sorting from 1.6 million replays would take too long and would overload the server. Couldnt the sorting be client-side?
No, it can't be. To sort by a property, you need all the records. For example, let's say you want to filter games to the map "| iCCup | Fighting Spirit 1.3", and sort the results by game length, descending. There are more than 160 thousand result games, how would you tell which are the 20 longest games without looking at each of them? That would require sending all 160 thousand games to the client, just to be able to sort and show the first 20...
When the data is limited, client side sorting can be implemented, and it is already done: look at player profiles for example, all tables are sortable by almost any column, on the client side. This is possible because all the data (table rows) are already on the client side.
|
On March 23 2023 20:52 Dakota_Fanning wrote:Show nested quote +On March 23 2023 15:18 MeSaber wrote:On March 22 2023 19:32 Dakota_Fanning wrote:On March 22 2023 16:17 MeSaber wrote:Rant: In Players tab, i would like to exclude players based on if they dont play BGH+Low, ie i want to find BGH or Low only players, maybe even be able to set a percent of played map type like 10%. Some have 1% or less of BGH/Low and i dont count those as BGH/Low players. In Games tab, i am trying to find players of 3v3 and 4v4 games AND only BGH/Low maps. Also i dont get any help how to use regexp map. If i could regexp the 'matchup' to include both 3v3 and 4v4 matchups like: XXXvXXX, XXXXvXXXX, ie wildcard a race and amount of players. Or 'format' be able to choose multiple options. Avg APM seems pretty useless, would rather have Avg EAPM. Avg APM can fluctuate a lot per player but EAPM usually does not. I would really like if you could use regexp in every field to find what ure looking for. This would require a regexp help-popup how regexp works ofc as most ppl dont know about it but its a strong tool. So my current confusion is how i find BGH/Low games with 3v3 or 4v4 format included. The lack of sorting is an issue too. It would be very useful. Edit: Be able to exclude short games like less than 5 minutes. Maybe repmastered should remove invalid games that are less than 2 minutes long (any game below 2 min is a disc). https://repmastered.app/?from=2021-01-01&mapMatch=substr&map=hunter&format=4v4&avgAPM=300 as an example. Thx for making this awesome website. Thanks for the feedback. I'd rather answer in general that addresses most of your issues / requests. Please keep in mind that performance is one of the first prio. If you have 10 replays, can you form any kind of query and sortig, the results will come in an instance. repmastered.app has more than 1.6 million replays as of now, and this number just keeps growing. To have "instant" results, the data has to be indexed in multiple ways to support queries and sorting. Building and maintaining indices is a significant task and space, and there are certain limitations. Performance (execution time) of queries and sorting will only be acceptable if they are backed by indices. Even though I would want to allow querying and sorting by all values, this is not possible in practice and is not possible due to database limitations. I carefully select filters and sort orders that seem useful / reasonable. I'll keep revisiting this issue from time-to-time, but I can't just allow all because without index support, filtering and sorting from 1.6 million replays would take too long and would overload the server. Couldnt the sorting be client-side? No, it can't be. To sort by a property, you need all the records. For example, let's say you want to filter games to the map "| iCCup | Fighting Spirit 1.3", and sort the results by game length, descending. There are more than 160 thousand result games, how would you tell which are the 20 longest games without looking at each of them? That would require sending all 160 thousand games to the client, just to be able to sort and show the first 20... When the data is limited, client side sorting can be implemented, and it is already done: look at player profiles for example, all tables are sortable by almost any column, on the client side. This is possible because all the data (table rows) are already on the client side.
Not saying that this is absolutely necessary by any means, nor do I think it is even all that useful of a thing to implement, and also I am coding-deficient. However, wouldn't a possible work-around be running these sorting algorithms say once per day, and then display the outcome of said sorting until the next time the sorting algorithm is done, instead of doing it on-demand per user? It's not ideal and I'm not sure how "expensive" this would be, but I feel like it might be a possible solution even if not ideal.
|
Dakota_Fanning
Hungary2345 Posts
On March 23 2023 22:44 Jealous wrote:Show nested quote +On March 23 2023 20:52 Dakota_Fanning wrote:On March 23 2023 15:18 MeSaber wrote:On March 22 2023 19:32 Dakota_Fanning wrote:On March 22 2023 16:17 MeSaber wrote:Rant: In Players tab, i would like to exclude players based on if they dont play BGH+Low, ie i want to find BGH or Low only players, maybe even be able to set a percent of played map type like 10%. Some have 1% or less of BGH/Low and i dont count those as BGH/Low players. In Games tab, i am trying to find players of 3v3 and 4v4 games AND only BGH/Low maps. Also i dont get any help how to use regexp map. If i could regexp the 'matchup' to include both 3v3 and 4v4 matchups like: XXXvXXX, XXXXvXXXX, ie wildcard a race and amount of players. Or 'format' be able to choose multiple options. Avg APM seems pretty useless, would rather have Avg EAPM. Avg APM can fluctuate a lot per player but EAPM usually does not. I would really like if you could use regexp in every field to find what ure looking for. This would require a regexp help-popup how regexp works ofc as most ppl dont know about it but its a strong tool. So my current confusion is how i find BGH/Low games with 3v3 or 4v4 format included. The lack of sorting is an issue too. It would be very useful. Edit: Be able to exclude short games like less than 5 minutes. Maybe repmastered should remove invalid games that are less than 2 minutes long (any game below 2 min is a disc). https://repmastered.app/?from=2021-01-01&mapMatch=substr&map=hunter&format=4v4&avgAPM=300 as an example. Thx for making this awesome website. Thanks for the feedback. I'd rather answer in general that addresses most of your issues / requests. Please keep in mind that performance is one of the first prio. If you have 10 replays, can you form any kind of query and sortig, the results will come in an instance. repmastered.app has more than 1.6 million replays as of now, and this number just keeps growing. To have "instant" results, the data has to be indexed in multiple ways to support queries and sorting. Building and maintaining indices is a significant task and space, and there are certain limitations. Performance (execution time) of queries and sorting will only be acceptable if they are backed by indices. Even though I would want to allow querying and sorting by all values, this is not possible in practice and is not possible due to database limitations. I carefully select filters and sort orders that seem useful / reasonable. I'll keep revisiting this issue from time-to-time, but I can't just allow all because without index support, filtering and sorting from 1.6 million replays would take too long and would overload the server. Couldnt the sorting be client-side? No, it can't be. To sort by a property, you need all the records. For example, let's say you want to filter games to the map "| iCCup | Fighting Spirit 1.3", and sort the results by game length, descending. There are more than 160 thousand result games, how would you tell which are the 20 longest games without looking at each of them? That would require sending all 160 thousand games to the client, just to be able to sort and show the first 20... When the data is limited, client side sorting can be implemented, and it is already done: look at player profiles for example, all tables are sortable by almost any column, on the client side. This is possible because all the data (table rows) are already on the client side. Not saying that this is absolutely necessary by any means, nor do I think it is even all that useful of a thing to implement, and also I am coding-deficient. However, wouldn't a possible work-around be running these sorting algorithms say once per day, and then display the outcome of said sorting until the next time the sorting algorithm is done, instead of doing it on-demand per user? It's not ideal and I'm not sure how "expensive" this would be, but I feel like it might be a possible solution even if not ideal.
For a specific filter / sorting setup it could be done, but that could also be covered with an index, so on-demand execution is also fast. Doing it for all possible filter / sorting combinations is again just too much and not worth imo.
|
Dakota_Fanning
Hungary2345 Posts
🎂 repmastered.app turns 3!!!
repmastered.app celebrates its third birthday 🎉! Since the second birthday:
450,000 new Replays uploaded (total: 1,733,333)
146 new 📅Events added (total: 746)
160,000 new 👨Players in the Database (total: 729,561)
1 million new 👬Player Connections (Unique Player Pairs) (total: 3,578,738)
2,300 new 🗺️Maps (total: 13,633)
370,000 new ▤Map Leaderboard Entries (Unique Map-Player Pairs) (total: 1,551,669)
Starting from 2001-01-01, on average:
There are 212.9 new replays every single day.
Or in other words: there's a new replay in every 6.8 minutes.
There is a new event in every 10.9 days.
To see exact and up-to-date numbers and even more stats, visit the ∑Stats page.
Thanks three Millions for the continuous replay contributions! 👏
Let's hope the 4th year brings even more replays! I encourage everyone to upload all your replays before they are gone forever. No registration is required. Become part of History, become part of StarCraft's Legacy!
|
Happy Birthday! Do we get cake now?
|
Dakota_Fanning
Hungary2345 Posts
This has come true yesterday:
The 📹Games page has a new filter: Include short games.
This is off by default, so games shorter than 110 seconds are not listed by default. These games are and were excluded from statistics.
You can list all games including short ones by enabling this filter.
|
Thats awesome! Thank you 
|
This database is huge! What's are the sources for non-user uploaded replays? I can find a lot old games (of what appear to be private replays) & so many ladder games of which I can't imagine they are all uploaded by users. It'd be nice to see date uploaded and the uploader/source.
Due to the sheer quantity a lot of the replays are never downloaded.. while there's some potential statistical value, the amount of bloat is enormous. Aside from some entertainment value some E-B rank reps might have the majority are low level games with (little to) no value.
Reducing this could solve a lot of future storage and traffic (and cost) issues.
|
Dakota_Fanning
Hungary2345 Posts
On August 23 2023 05:45 Smorrie wrote:
This database is huge! What's are the sources for non-user uploaded replays? I can find a lot old games (of what appear to be private replays) & so many ladder games of which I can't imagine they are all uploaded by users. It'd be nice to see date uploaded and the uploader/source.
Due to the sheer quantity a lot of the replays are never downloaded.. while there's some potential statistical value, the amount of bloat is enormous. Aside from some entertainment value some E-B rank reps might have the majority are low level games with (little to) no value.
Reducing this could solve a lot of future storage and traffic (and cost) issues.
Source of non-user uploaded replays is the internet and Blizzard ladder. The identity of the uploader is private info. The upload timestamp is published, just hidden by default. Click on the ☷ Columns link and check Uploaded. This will display the upload date, toolip also shows the exact timestamp.
And yes, since there are so many replays (10+ millions!), most will never be downloaded but contribute invaluably to all kinds of statistics (matchup, map, player etc.). They are also invaluable to the players themselves, for example you probably don't care about my replays and my player stats, but that's what I care about the most.
|
Thanks for the very quick response!
While it's awesome there is so much data I do have mixed feelings about the gathering methods, especially since there's no way to opt-in or opt-out for players. While it's true I'm not interested in other players' statistics, I'm not sure if I'm interested in my own either - especially when the data set is limited (when I haven't uploaded my own replay folder). I also think for transparency it would be good to show the uploader.
That aside, great work on this platform, looks really solid.
Has any statistical data been released before? Would love to see some if available.
It would be awesome to find things such as map balance within different mmr ranges.
On a more detailed level you could also extract build order effectiveness. This would of course require for build orders to be categorized within a variety in conditions, but it should be very possible to add labels such as '3 hatch muta', '1 rax FE SK terran', etc. Which in turn would also give the opportunity to search by build order. This feels like a huge undertaking without community support though (& also data processing).
Also awesome that you're still actively maintaining the platform and adding features! Cheers
|
On August 24 2023 02:47 Smorrie wrote:
Thanks for the very quick response!
While it's awesome there is so much data I do have mixed feelings about the gathering methods, especially since there's no way to opt-in or opt-out for players. While it's true I'm not interested in other players' statistics, I'm not sure if I'm interested in my own either - especially when the data set is limited (when I haven't uploaded my own replay folder). I also think for transparency it would be good to show the uploader.
That aside, great work on this platform, looks really solid.
Has any statistical data been released before? Would love to see some if available.
It would be awesome to find things such as map balance within different mmr ranges.
On a more detailed level you could also extract build order effectiveness. This would of course require for build orders to be categorized within a variety in conditions, but it should be very possible to add labels such as '3 hatch muta', '1 rax FE SK terran', etc. Which in turn would also give the opportunity to search by build order. This feels like a huge undertaking without community support though (& also data processing).
Also awesome that you're still actively maintaining the platform and adding features! Cheers
https://tl.net/forum/brood-war/572031-ladder-statistics-race-distribution-2021
Like this?
|
Dakota_Fanning
Hungary2345 Posts
On August 24 2023 02:47 Smorrie wrote:
Thanks for the very quick response!
While it's awesome there is so much data I do have mixed feelings about the gathering methods, especially since there's no way to opt-in or opt-out for players. While it's true I'm not interested in other players' statistics, I'm not sure if I'm interested in my own either - especially when the data set is limited (when I haven't uploaded my own replay folder). I also think for transparency it would be good to show the uploader.
That aside, great work on this platform, looks really solid.
Has any statistical data been released before? Would love to see some if available.
It would be awesome to find things such as map balance within different mmr ranges.
On a more detailed level you could also extract build order effectiveness. This would of course require for build orders to be categorized within a variety in conditions, but it should be very possible to add labels such as '3 hatch muta', '1 rax FE SK terran', etc. Which in turn would also give the opportunity to search by build order. This feels like a huge undertaking without community support though (& also data processing).
Also awesome that you're still actively maintaining the platform and adding features! Cheers
I'm not gathering replays by hacking servers or users' computers. I'm gathering replays that are publicly available on different sites and forums, posted on public discord servers etc.
My point is this: if I wouldn't collect and publish these, that doesn't mean these replays are not accessible on the internet and no one else does this, maybe secretly. This also goes for the opt in and out. If you could opt out, that doesn't mean the data isn't collected or someone else doesn't collect it either. This way at least you know what's available on the internet.
Think of it this way: you play a 1v1 game vs someone, but you want to keep the game private. You can, of course, keep your replay private, but you can't keep the opponent's replay private: if he wants to, he can publish his replay and there's nothing you can do to keep the game a secret.
I'm not sure / convinced the replay uploader should be public info. Do note however that when the replay saver can be identified / deduced from the data saved in replays, that is displayed on the Game details page.
Current gathered and published stats are availabe on this page: https://repmastered.app/stats These stats are updated "real time" as new replays are uploaded. This page is semi-interactive: you can apply certain filters like map class, ladder and years. There are also stats gathered per map, per player etc, which are displayed on map, player etc profile pages.
Build order extraction and related stats are non-trivial and less accurate to say the least, because replays store limited information (it's not recorded which buildings or units complete, only the commands / actions issued by players are recorded).
More statistics of any kind of course are possible in the future, the first step is always the data gathering.
|
Thanks for sharing your perspective. I fully understand, since the platform needs to be fed with replays
+ Show Spoiler [just some food for thought] +I don't really agree, but I don't want to litter the thread with negative sentiment since I do think the website itself is great. I'll share my thoughts and then leave it at.
While replays might be publicly accessible, it doesn't necessarily mean the players, or the original host, consents to having them republished. Likely, for many cases, players aren't even aware or would like them to be published in the first place.
While replay 'ownership' always involves at least 2 parties, using this as an argument to rehost the replays isn't really genuine either. In practice, for majority of the ladder replays neither player has intended (nor are they aware) that the replays are being published, mainly due to a backdoor into Blizzard's ladder replay library.
In that sense it's a little hypocritical to not share the uploader's name, even thought this doesn't really matter too much anyway since anyone can create a random login to anonymously upload replays.
Of course everything is in good faith and I'm not contesting any legal ownership or whatsoever. It's just the principle - especially in this day and age where digital privacy on a wider scope involves a lot of issues. It appears there is very little (or no) consideration for the players & everything is kept anonymous.
Not requesting any changes, just providing some food for thought.
In the end, the community is small and I do believe that any development and/or efforts made to keep it alive are always welcome. Additionally, I think this makes any other replay analyzer obsolete.
|
Dakota_Fanning
Hungary2345 Posts
Summary of recent updates:
⚔️Matchmaking Info!⚔️
Matchmaking info for Blizzard ladder games is now available!
This information is not obtained from replays, and may be acquired / updated later, in the background.
Added an MM Info column on the 📹Games page which displays ladder season, gateway and average MMR with rank of the game. This info is also added to Game details.
Added an MMR column on the Players tab of Game details.
This column displays the players' initial MMR (with rank) before, and MMR change after the game.
⚔️Matchmaking Activity and Trends in Player Profiles!⚔️
Added a new MMRs Min·Latest·Max column to the 👨Players page.
This column displays the overall minimum, latest known and maximum MMR of players, with ranks.
Players can now be sorted by Latest MMR and Max MMR on the 👨Players page.
Added Matchmaking Activity and Trends to player profiles.
This visualizes the player's min/last/max MMR based on matchmaking seasons. Detailed stats can also be displayed (it is collapsed by default).
Added Matchmaking Gateway stats to player profiles.
This section shows the player's stats (including min/latest/max MMR) based on gateways.
Miscellaneous changes and new features
The 📹Games page has a new filter: Include short games.
This is off by default, so games shorter than 110 seconds are not listed by default. These games are and were excluded from statistics.
You can list all games including short ones by enabling this filter.
Replaced the global "Stretch tables" overlay option with a new Tables formatting listbox. It has the following options:
- No wrap: This option will not break text content of table cells, so all rows will have equal hight. Horizontal scrollbar will appear if the table doesn't fit on the page.
- Wrap: This is the old, default behavior: Text content will break into multiple lines if otherwise the table would not fit on the page.
- Stretch: This is the old "stetched" behavior: tables will stretch to full page width if they would be smaller.
Highlight game results: When you filter games by players on the 📹Games page, game results will now be highlighted based on the result of the searched players.
If you filter games to a single player (or multiple aliases of a player), the table will be colored based on when that player has won or lost.
When filtering to multiple players, the coloring is only applied if all the listed players are on the same team.
Changed 👨Players, 🗺️Maps, 👬Playmates, ▤Leaderboards and 🆚Machups: replaced the default "Total Games" column with "Included Games", and they are now sorted by included games (instead of total games). Using the ☷ Columns link you can still turn on/off the Total Games, Included Games and Excluded Games columns.
The Years stats on the ∑Stats now only display the number of included games (instead of total games).
Cosmetic updates:
- Map class, ladder and race values are now colored eveywhere.
- Matchups are also colored. Matchups have a race distribution color background calculated from the races of the matchup.
- Race and matchup stats table cells inside Year Stats on the ∑Stats page now also use race colors.
View the complete and up-to-date list on the 📰News page.
|
Dakota_Fanning
Hungary2345 Posts
Also one unfortunate but important thing.
repmastered.app now hosts well over 10 MILLION replays! Operational costs have increased significantly. I hate begging, but to keep the site running, your financial help is needed.
I'm not asking for a profit nor payment for my work. I work on this for free. But I do have to pay a considerable amount for the server resources every month, and that I have also done voluntarily, from my own pocket, for years now, and I have done it to give something to the community. But it has grown to a considerable amount which I may not willing to pay going forward, which would require to shut down the site.
So if you like the site and think it gives value to the community, consider donating to keep the site live. Thanks.
|
Thank you for your hard work and for these awesome updates!
|
On September 07 2023 20:40 Dakota_Fanning wrote:Also one unfortunate but important thing. repmastered.app now hosts well over 10 MILLION replays! Operational costs have increased significantly. I hate begging, but to keep the site running, your financial help is needed. I'm not asking for a profit nor payment for my work. I work on this for free. But I do have to pay a considerable amount for the server resources every month, and that I have also done voluntarily, from my own pocket, for years now, and I have done it to give something to the community. But it has grown to a considerable amount which I may not willing to pay going forward, which would require to shut down the site. So if you like the site and think it gives value to the community, consider donating to keep the site live. Thanks.
Great work and thanks for repmastered and especially for the recent updates!
Im Wondering how is it possible that the total replay count has grown like 300% in 2023?
|
Dakota_Fanning
Hungary2345 Posts
On September 07 2023 23:38 earob84 wrote:Show nested quote +On September 07 2023 20:40 Dakota_Fanning wrote:Also one unfortunate but important thing. repmastered.app now hosts well over 10 MILLION replays! Operational costs have increased significantly. I hate begging, but to keep the site running, your financial help is needed. I'm not asking for a profit nor payment for my work. I work on this for free. But I do have to pay a considerable amount for the server resources every month, and that I have also done voluntarily, from my own pocket, for years now, and I have done it to give something to the community. But it has grown to a considerable amount which I may not willing to pay going forward, which would require to shut down the site. So if you like the site and think it gives value to the community, consider donating to keep the site live. Thanks. Great work and thanks for repmastered and especially for the recent updates! Im Wondering how is it possible that the total replay count has grown like 300% in 2023?
The majority of those replays are Blizzard ladder replays. User replay uploads also increased significantly (either anonymous or registered users).
|
i appreciate the time/work you put into this site
but that being said
1) you should provide a link to the original website where the download was scraped (ygosu, yaoyuan, reps.ru, etc)
2) especially if you are going to charge for downloads due to "increasing abuse"
3) or, how about a simple captcha to prevent unwanted downloads? im not going to pay USD for replays for a 25 year old game
thanks
|
|
|
|
|
|
EPT
