Tired off not knowing the GSP required for Elite? I have a solution, but I need your help!

[Hotrod08]

Hello guys,

This Friday is the official release of EliteGSP.com (also EliteSmash.com), a community-run SSBU website created to provide you the best ballpark estimates of the GSP needed to earn Elite Smash!

Before Friday's release though, I need your help getting accurate data! If you have obtained Elite Smash within the last 3 days, or earn Elite in the future, please go to the site and post your GSP(s), the number right when you earned Elite, so that the site can have the most accurate data!

Haven't earned Elite yet? Don't fret! (I'm right there with you >.<) Starting Friday, you'll be able to see what GSP you need to earn for each character in the game! Just make sure you enter your GSP once you finally hit Elite

Thank you all again, and if you have any questions, feel free to tag me below and I'd be happy to help!

Dylan (Hotrod08)

Link to preview of Friday's launch site

 

[Leoinu]

GSP needed for elite smash constantly changes depending on everyone elses score. There is no specific thresh-hold you need to pass, you just have to be in the top percentiles of GSP. So unless youre constantly getting data from Nintendo’s servers, your site is going to produce unreliable information.

 

[Oneiros5321]

GSP needed for elite smash constantly changes depending on everyone elses score. There is no specific thresh-hold you need to pass, you just have to be in the top percentiles of GSP. So unless youre constantly getting data from Nintendo’s servers, your site is going to produce unreliable information.

This ^^^

Unless you are aware at any time of how many new players joined the game, there is no way to have accurate informations. The GSP needed to reach Elite Smash is constantly changing.

Also, I think that would make even more people obsessed with their GSP which I don't want to encourage =')

 

[Hotrod08]

GSP needed for elite smash constantly changes depending on everyone elses score. There is no specific thresh-hold you need to pass, you just have to be in the top percentiles of GSP. So unless youre constantly getting data from Nintendo’s servers, your site is going to produce unreliable information.

Yes, the GSP will be constantly changing, which is why the website only uses the last week of GSPs entered when it does its math. As the site launches, there will still be the opportunity for people to constantly post what GSP it took for them to reach Elite Status, thereby always keeping the GSP updated (and increasing)

 

[UltimateXsniper]

I definitely can see the use for this. I don't entirely see why the others would dismiss this. Like, it's already given that elite rises as people are joining in online and I can also assume that the site will be updated regularly for that. As long as people enter what they got in recently with elite, and the site can make a good enough calculation based off of those recent posts, then I think this would be reliable enough to make an rough estimate or a good idea on where you should be to get into elite.

However, I can see issues coming from this is that people could falsely put in information and if enough people can keep it running. I think getting the attention may be the hardest part to this really as you're trying to gather data from people pretty much constantly. That's definitely going to be hard especially since you're also gathering information with characters they got in with.

I like the idea and it is an interesting project, there's potential to it, but it's up in the air whether or not you can get a lot of people going for it. I'd suggest showing your site to other media if you haven't already and hope people can play along with it.

 

[Lord_Ghirahim]

I think it might depend on region? Only basing this theory on the fact that I currently need 3.2 mil for Elite but I see streamers in the US getting characters into Elite with under 3 mil.

 

[Hotrod08]

I definitely can see the use for this. I don't entirely see why the others would dismiss this. Like, it's already given that elite rises as people are joining in online and I can also assume that the site will be updated regularly for that. As long as people enter what they got in recently with elite, and the site can make a good enough calculation based off of those recent posts, then I think this would be reliable enough to make an rough estimate or a good idea on where you should be to get into elite.

However, I can see issues coming from this is that people could falsely put in information and if enough people can keep it running. I think getting the attention may be the hardest part to this really as you're trying to gather data from people pretty much constantly. That's definitely going to be hard especially since you're also gathering information with characters they got in with.

I like the idea and it is an interesting project, there's potential to it, but it's up in the air whether or not you can get a lot of people going for it. I'd suggest showing your site to other media if you haven't already and hope people can play along with it.

Well, with the release of the website on Friday and getting prerelease data, I currently have had 257 entries on the website, so far so good on data. I also was able to talk to one of the Smash Bros subreddits with 100k subscribers and they plan on keeping the website stickied on their subreddit for a bit, which will really help lift off the website at the beginning. Finally, I have some measure in place on the website to prevent people from abusing posting numbers too many times or too high/too low, so we'll see how that plays out.

Overall, I'm pretty estatic for the site! I think it has a good chance of succeeding

 

[Leoinu]

Yes, the GSP will be constantly changing, which is why the website only uses the last week of GSPs entered when it does its math. As the site launches, there will still be the opportunity for people to constantly post what GSP it took for them to reach Elite Status, thereby always keeping the GSP updated (and increasing)

So you're not only hoping randoms input their GSP, which constantly switches daily, sometimes by the hour. You're also hoping they come back every week to add info to a site that will no doubt have inaccurate information? Cut your loses. Its a good idea with a horrible environment to insert it into.

Well, with the release of the website on Friday and getting prerelease data, I currently have had 257 entries on the website, so far so good on data. I also was able to talk to one of the Smash Bros subreddits with 100k subscribers and they plan on keeping the website stickied on their subreddit for a bit, which will really help lift off the website at the beginning. Finally, I have some measure in place on the website to prevent people from abusing posting numbers too many times or too high/too low, so we'll see how that plays out.

Overall, I'm pretty estatic for the site! I think it has a good chance of succeeding

257? Out of literal millions? bruh....

 

[Oneiros5321]

257? Out of literal millions? bruh....

To be fair he only needs to know the GSP of Elite Smash players to be accurate. Which is definitely not "millions" of people.

 

[Leoinu]

To be fair he only needs to know the GSP of Elite Smash players to be accurate. Which is definitely not "millions" of people.

Fair point. So lemme change it a bit. Lets be generous and say like the top 1% of players with a high enough GSP are in Elite smash, because we've been told its based on the top percentile. 1% of 3 million is still 300,000. So yes, 257 for a theoretical 300,000 is still ridiculously small and will produce no reliable data.

Hell to make matters worse. Thats a number reported spanning over 74 playable fighters. I wont be surprised if a number of them will either have 0 reports, or less than 5.

The data produced from this tool will be completely unreliable and will only serve to make people more pissed over the hell-system of GSP.

Edit: Hell to break the concept even further, how many players fall out of Elite smash? Lets say I did want to use your tool and came back every week to update with information. "Yeah, still Elite Smash with Falco.....I dont know what to do now." So not only do you have a moving target with very little data, it only gets wrose as the number of entered numbers decrease as people will either:

1) Not come back
2) Legit dont know what GSP is needed to be Elite with a certain character, so they CANT enter in new information.

No matter what, your amount of reports to base an average on will only get smaller and smaller. As if it couldnt get worse than what I've mentioned already.

 

[UltimateXsniper]

Fair point. So lemme change it a bit. Lets be generous and say like the top 1% of players with a high enough GSP are in Elite smash, because we've been told its based on the top percentile. 1% of 3 million is still 300,000. So yes, 257 for a theoretical 300,000 is still ridiculously small and will produce no reliable data.

Hell to make matters worse. Thats a number reported spanning over 69 fighters. I wont be surprised if a number of them will either have 0 reports, or less than 5.

The data produced from this tool will be completely unreliable and will only serve to make people more pissed over the hell-system of GSP.

Well you still wouldn't need those 300,000 players in elite either. After all it's about what GSP is needed to get in elite. People just need to report what GSP they just got into elite with and you could already get a decent idea on what GSP you'll need after several people posting similar results in a short notice.

I don't think the numbers of players to that scale is an issue itself but more about the timing as you're right with the characters. With a bunch of characters you'll need at least several people for each character at least a few times a day to roughly guess where the GSP could be down to the ten thousands range.

 

[Hotrod08]

1% of 3 million is still 300,000.

1% of 3 million is 30,000.

The goal of the website is to give players a ballpark GSP required to enter Elite, not be 100% accurate. No method that's not Nintendo itself will give 100% accurate numbers, but if the website displays a GSP that's only 20,000 GSP off of being the true GSP needed to get into Elite, I'd be more than happy.

The website also only uses GSPs that are within 1 week old or newer, throwing out old data and using only new data as it's coming in.

 

[Leoinu]

1% of 3 million is 30,000.

The one time I **** up and misread a calculator. Rip lol

As for the ballpark statement. I knew that since the start. What I’m concerned is with such little reports, said ballpark could be massive and still unreliable. Best of luck I guess. Within 20,000 estimates would be successful imo as well. Since thats basically 1 or 2 wins

 

[Veggieburger_802]

This is a brilliant idea, Hotrod. You've done a great job putting together a tool that I'm sure will be very useful for lots of people in years to come. If you get enough data, I imagine you could implement a weighting algorithm to account for the drifting of thresholds, which could generate even more accurate numbers.

For now, though, I have a suggestion that I think might make the tool more valuable: a little note under each character's GSP that indicates how many data points are being averaged to produce that number. So, for example, if only one Elite Smasher has submitted their GSP for a certain character in the past week, players of that character will know to take the number with a grain of salt. Conversely, if a player sees that a character's average Elite GSP has been calculated based on dozens or even hundreds of submissions from the past week, then he can be pretty confident in the number's accuracy.

But this is just an idea. I'm not professing to understand how your system works, or asserting that I could have done a better job myself. I think you've created something very cool, and I'm excited to see it grow. Best of luck.

 

[leafgreen386]

1% of 3 million is 30,000.

The goal of the website is to give players a ballpark GSP required to enter Elite, not be 100% accurate. No method that's not Nintendo itself will give 100% accurate numbers, but if the website displays a GSP that's only 20,000 GSP off of being the true GSP needed to get into Elite, I'd be more than happy.

The website also only uses GSPs that are within 1 week old or newer, throwing out old data and using only new data as it's coming in.

One week is a *very* long time when it comes to GSP. The elite smash threshold could easily change by 200-300k in that time. You're going to end up grossly underestimating if you use data from that long ago.

The most accurate way to find elite smash GSP, within a few thousand, is to multiply a top ranking player's GSP by 0.965. The top ranking players are all around 3.32 mil right now, which should give an ES threshold of about 3.2 mil.

edit: It could also be 0.97, which would place the threshold at 3.22 mil. The threshold is either top 3% or top 3.5%, and while we collected data at one point that suggested it's 3.5%, it's a very slim difference that a minor difference in reporting time can throw off.

 

[Hotrod08]

This is a brilliant idea, Hotrod. You've done a great job putting together a tool that I'm sure will be very useful for lots of people in years to come. If you get enough data, I imagine you could implement a weighting algorithm to account for the drifting of thresholds, which could generate even more accurate numbers.

For now, though, I have a suggestion that I think might make the tool more valuable: a little note under each character's GSP that indicates how many data points are being averaged to produce that number. So, for example, if only one Elite Smasher has submitted their GSP for a certain character in the past week, players of that character will know to take the number with a grain of salt. Conversely, if a player sees that a character's average Elite GSP has been calculated based on dozens or even hundreds of submissions from the past week, then he can be pretty confident in the number's accuracy.

But this is just an idea. I'm not professing to understand how your system works, or asserting that I could have done a better job myself. I think you've created something very cool, and I'm excited to see it grow. Best of luck.

I thought of that as I was developing the side, but there's word going around that the GSP limit is the same for all characters within the game, so any character-specific GSPs might or might not be relevant. At least for now though, you can go to the bottom of the page and see how many entries have been used in the last week of calculations.

Edit: Thanks for the complement! I'm glad you enjoyed the site

 

[Hotrod08]

One week is a *very* long time when it comes to GSP. The elite smash threshold could easily change by 200-300k in that time. You're going to end up grossly underestimating if you use data from that long ago.

The most accurate way to find elite smash GSP, within a few thousand, is to multiply a top ranking player's GSP by 0.965. The top ranking players are all around 3.32 mil right now, which should give an ES threshold of about 3.2 mil.

edit: It could also be 0.97, which would place the threshold at 3.22 mil. The threshold is either top 3% or top 3.5%, and while we collected data at one point that suggested it's 3.5%, it's a very slim difference that a minor difference in reporting time can throw off.

A week is rather long, yes, and I might shorten that number up depending on how consistently I'm getting GSP entries on the website. If I'm getting enough data, then I might be able to shorten it to 3 days instead, making it that much more accurate.

I'd prefer to use people's data, rather than taking in top player's GSPs, as you really can't tell if they're truly top of the top currently, and there's no proof what the % is of the top players yet. Also, what system would there be in place to report those top players GSPs? It would get rather messy fast, along with exhausting on my part since I'd have to constantly monitor it as opposed to the automatic system I have in place currently.

 

[Lions’s Sin of Pride]

I keep hearing different things for people and I’ve also seen different scores for people who are in Elite.

But for me 3.2Million is what I need to get my characters into Elite.

 

[leafgreen386]

A week is rather long, yes, and I might shorten that number up depending on how consistently I'm getting GSP entries on the website. If I'm getting enough data, then I might be able to shorten it to 3 days instead, making it that much more accurate.

I'd prefer to use people's data, rather than taking in top player's GSPs, as you really can't tell if they're truly top of the top currently, and there's no proof what the % is of the top players yet. Also, what system would there be in place to report those top players GSPs? It would get rather messy fast, along with exhausting on my part since I'd have to constantly monitor it as opposed to the automatic system I have in place currently.

Players like Leffen, ZeRo, etc, will always be top of the top. Even if they're not #1 currently, they don't need to be, since even if they were a few thousand off, it would barely affect the calculation.

Just a quick look at their stream once in between matches is good enough, since it'll show their highest character's GSP regardless of who they're playing. You could probably even set up a bot to do it.

 

[Hotrod08]

It's an idea if I can't ever make the current website accurate enough to work, but I'm relatively content with the numbers it's pulling so far, only about 1.2% under what's been recently posted as the Elite Smash limit.

 

[Telek]

If i'm not mistaken, the "X hours ago" on twitch videos is how many hours ago the broadcast started, correct? If so, it looks like ZeRo was at 3307 about 8 hours ago. Around 2 hours later, I had the elite smash between 3,186 and 3,217, which puts elite smash between the top 4.1% and top 3.2% (roughly, though each of those % could conceivably be slightly higher w/ the time difference). In any case, I honestly believe that you can throw out most of the data points being entered on your website and just keep the current highest GSP recorded not in elite. That's the real bellwether of where the line is, because once you're above the elite threshold you can't know where the line is until you fall back below (which can happen if left alone near the line for long enough). Importantly, that information never ever goes stale because that number will never be above the threshold as the line drifts, since it goes in the opposite direction, so you're always updating outdated versions of it with better estimates. I've been using this the past week or so by getting a couple of fighters to within a few thousand of each other right near the cutoff, and letting one of them fall out of elite naturally as people pass them. The closer we can get an example of two people with one below and one above the line, the more accurately we can track the number.

 

[peekpeek]

I keep hearing different things for people and I’ve also seen different scores for people who are in Elite.

But for me 3.2Million is what I need to get my characters into Elite.

It goes up by 20-50k per day as the size of the overall player pool increases.

 

[Lions’s Sin of Pride]

It goes up by 20-50k per day as the size of the overall player pool increases.

Our personal GSP or the overall GSP?

 

[Hotrod08]

He's talking about the max GSP, but your own increases proportionally with the max as well

 

[leafgreen386]

Our personal GSP or the overall GSP?

Yes.

 

[Telek]

My donkey fell out, which is perfect, current GSP threshold is between 3,237,020 and 3,239,044