Updating the BiggerPockets Influence System
Joshua D.Pro Member
BiggerPockets Founder · HI · Member since 2008 · 16k+ posts · 5k+ votesHey All -
One of our upcoming projects is to update our influence system to one that better reflects a user's influence on the site.
We will be transforming the system from the current limitless influence scale to a 0 to 100 scale, where all new users start at 50, and can work down or up from there. Active users who have influence points currently will be placed in the new system based on their current influence.
That said, one of the important factors of our new system will be a time decay of influence, so those people who may have been active in the past, but who are no longer active, will slowly lose points over time.
Ultimately, we're interested in hearing from you, what makes you trust OR NOT trust someone on the site?
We want to take as many factors into account as possible so this system has some meaning and shows at any given time who the influencers are, and I'm looking to you for your feedback.
One of the things I'm most interested in hearing about is user feedback on negatives and the drivers of a user losing "points" on our scale.
Thanks and I look forward to a lively and respectful debate.
After reading all the comments thus far, I am leaning towards this:
Screw it, delete the influence system all together, who cares. You know who you like and respect after reading their posts so just go with that. If you are new, it will become abundantly clear who the main contributors are, who is wise in what areas and who is not.
The influence system would influence me more if it were easier to understand. i also think it and t be beneficial to make the rules more obvious, so that people new to the site are easier able to find out they are about to violate the rules. I say this because I violated the rules without knowing it or intending to do so. Perhaps it was my lack of attention, but hey, we're all busy, and reading a website's policy posting wouldn't find it's way to the top of my to do list.
My biggest concern, however, is finding a way to expose those people who are representing as fine upstanding professionals, but who in fact have been found guilty by the court system of being anything but. An example would be a disbarred real estate attorney who was disbarred for misappropriation of client funds. Twice.
Anyone who googled the person could easily find that out, but the influence system doesn't take that into consideration. Not sure how it could. Maybe a link in each person's profile that links to a google search for that person. Of course it would only work if the person gave their full name.
Maybe there is no solution to this, but it has bothered me since I once saw someone recommending the above-referenced person as reputable just because he was an active poster. So I'm raising the issue.
Maybe instead of the vote system that is currently in place a question of was this post helpful to you and use the number of yes and no responses to factor into the score.
Being new here as well, I would like the, Was it helpful, yes or no, idea a lot, as you would get great feedback and possibly even work out the system where you could sort by Most Helpful, which would be wonderful in the topics that have more than 4 pages of responses. Then instead of awards or badges, you could simply have Member Name, Number of Posts, and Number of Helpful Posts. I wouldn't worry about marking or advertising the No votes as you guys all seem to be able to take care of those guys very well here, but you could use it as an administration tool to find out easily if someone needs to be monitored for problems.
Having spent ridiculous numbers of hours debating this topic when I worked at eBay -- which has a horrible influence system but one of the largest on the web -- here are two things to consider, both of which I'm sure Josh is aware of:
1. There is no perfect system, the more complex it becomes, the harder it is for the typical user to manipulate but the easier it is for the determined user to manipulate;
2. The goal of influence for the owner of the site is often at odds with the goal of influence for users of the site. In the case of BP, Josh would probably love to use influence to encourage discussion by all members (increasing the value of the site), while users want it to identify reputable contributors (predicating the value of the site on just a few very influential, users).
That said, here's my suggestion (and my rationale behind it):
- Influence is only gained by getting upvoted on a post, not by starting threads, not by getting responses to threads, etc. Purely by someone indicating that they thought your post was good.
- The only real way to game the system this way is to have people collude (explicitly or implicitly) by upvoting each other's posts. This can be mostly mitigated by assigning influence on a progressive scale as opposed to a linear scale (i.e., valuing higher number of votes in a single post higher than lower number of votes).
- As an example, if I post something, I could get influence points on the following scale (this is based on algorithm that I think works well -- let me know if you want it, Josh):
- 0 Votes = 0 Influence
- 1 Votes = 0 Influence
- 2 Votes = 1 Influence
- 3 Votes = 2 Influence
- 4 Votes = 2 Influence
- 5 Votes = 5 Influence
- 6 Votes = 6 Influence
- 7 Votes = 7 Influence
- 8 Votes = 8 Influence
- 9 Votes = 14 Influence
- 10 Votes = 15 Influence
While this won't eliminate collusion, it will dampen the effects of it and likely not make it worthwhile to pursue for most/all contributors. And, if some contributors do find it worthwhile to try to collude, this system should make it easily apparent to anyone who has analytic data who the perpetrators are, making it easy to manually mitigate through whatever means are deemed necessary.
- Btw, influence systems are much like security systems -- the real key to mitigating a breach is policy, not automation. In other words, make sure everyone knows when it's appropriate to upvote and when it's not appropriate. And when it comes to building influence, in my opinion, the reasons to upvote should be purely based on quality of content (not popularity, not agreement, etc).
- The other benefit of this approach is that it should be pretty easy to retroactively assign influence based on historical data from the site.
- The voting system on posts needs to be encouraged. Use will beget use. Reiterate this to users at every opportunity.
- Lastly, I'm not a fan of time-decay in influence. This hurts the site owner (discouraging people from coming back if they've left for a while for whatever reason) and hurts the users (people start to look less valuable just because they haven't been around, which is an inaccurate perception).
My first choice would be to eliminate it completely as suggested by others. Anyone really interested in RE and BP can research particular posters and their threads to determine if researcher wants to follow them. There are some members that have tons of posts/influence and few care to follow them, due to BS or expertise in areas the researcher doesn't care about.
In second place, imo (and far back) would be JScott post. Again, someone could build up influence in areas that researcher/newbie has little interest in. If there was some way to tie J Scott suggestion in with a method to determine "expert listing" of poster in certain areas, that might work but would be difficult. It could replace the "awards" system currently in place which serves little purpose, imo.
Rich
Interesting Topic.
As a sporadic poster yet avid reader, I tend not to look at influence. Even when I first got here, to me it was more a measure of longevity rather than that of authority. As an interesting observation, it seems that with a sufficiently large post count, it begins to normalize around 10:1 being that influence tends to hover around 10x the post count.
Geeky statistics aside, I would give two opinions.
On the simple side of things, get rid of it. It doesn't seem to hold a measurable value in most cases and eliminating it would seem to be the easy fix. However, on that note, I do like seeing individual posts voted up. Whether that tied into a larger aggregate value or not would be something else - maybe it could be tied to some form of achievement rather than a generic influence.
On the more complex side, I do think there's significant value in subject matter influence. Some here are very, very knowledgable in specific aspects of the business. They should be given some form of at-a-glance recognition so a newbie knows to possibly hold their opinion at a higher level than some others. This may be forum specific since topic or thread specific would probably be a nightmare to program effectively. But in that case, specific subject matter influence would hold weight.
In thinking about it a little more, I think one of the problems is that it does seem to have a fuzzy definition. For someone just coming in, what is good influence? Right now, the system only shows a number, but there's no context to it. There is no threshold for judging. The system has an implication that in order to know what is good influence a person has to have some experience here, which by then probably negates the need for influence in the first place since they know who are the A+ posters and who aren't.
I'm completely new but have been lurking a couple of weeks. That being said, having run some forums in the past myself, I would suggest simplifying the system or deleting it all together.
To simplify it, I would suggest a range of +10 to -10. positives are good, negatives are bad. This is intuitive. Also, numbers higher than 10 get lost on some people when glancing, so don't exceed the number of fingers people have if you want it to be intuitive and simple without explanation or instructions.
I first thought this when there was only 1 or 2 replies to this thread but wasn't going to say it and sound like a downer, but I agree with Will, just get rid of it. Sounds like the whole thing is a headache. I could care less how many influence points someone has.
I am even more so of the opinion that it should be deleted all together.
One idea I do like is to simply have a badge system and in order to get final approval of that badge, Josh (and perhaps his selected few) would have to approve/agree with it.
One of the things many feel is missing is who they should hold more weight with on topics. For instance, if a question regarding rehabbing was asked, and J Scott & myself gave answers, along with Jon Holdman (I am only using us three as examples because we are all well known) a new member would have the ability to see a badge on J's profile stating something like "rehab expert" as well as mine, and Jon's would have something like "buy & hold expert" and/or "SDIRA expert". This is not to say that Jon's answer would be incorrect (it would likely not be), just to show others who in particular have extensive experience in a specific strategy.
How each member would be able to earn each badge - I do not know, perhaps a lot of the long time members who already are established, it would be easy for them to be assigned badges by Josh and company, while those that are not, I have no answer at this point.
I second Will kill the influence and add different badges based on performance not blogs. I see how blogs are valuable but if it was readily apparent that a member had a blog say in their signature I think that commands it's own influence / respect depending on the blog content.
2nd suggestion follow ... add categories. User defined. Ex: I could go and type teachers, deal contacts, I want to check back on them ... then I populate that with people. It would help newer member like me keep aware of why that person is on the follow list which is great when you are getting to know everybody.
I've found little value in the current point system. Like some have said once you've been around awhile you learn who gives serious thought to what they say and gives to the community.
To me what matters is what does the community want these influence posts to represent?
To me quantity doesn't mean very much. Finding a way to measure quality means more IMHO. Not sure how that might be accomplished.
A long winded post may not necessarily mean it has greater quality. The number of posts may indicate that at least some were of value, but not necessarily all or any of them.
It will be interesting to hear some of the ideas that come out of this.
With that said, lets just get rid of it all together and go with the badge system as I offered before. If you are a top rehabber and contributor to the site, you should have a "expert rehabber" badge and so on for all other RE investment strategies. Forget points, and badges have nothing to do with number of posts, but quality of the expertise of that member. This would require such approvals from Josh and company so if you were looking for ideas that do not add to your plate Josh, good luck with that. :)
As a long time poster and high post count, my ratio is only 7.5 to 1. While reading this comment then looking at many others, I have not found that average to be true.
...
Will,
In your case, the influence points did not start with your first post - you had a few thousand initial posts prior to the influence scoring system's start, and those initial posts did not get any credit for influence with the model that was put into place. If you had been given some credit for those few thousand posts, your influence number would probably rank you much higher. Ditto for Jon Holdman and Josh Dorkin when it comes to post counts and why influence scoring isn't much higher than it is.
Has anyone here read the current Influence point system lately? It's all games to keep people hungry to get them to the top....of what? I don't know. 40 points for getting somebody to sign up? Points for being a "narc" and reporting "violations". J Scott said earlier in this thread that what a community needs and a site owner needs are at odds. He's right.
I say: Josh...be brave, be brave! Build it without games and without fake approvals and fake awards and they will come! Focus on making the site REALLY easy to use. It it needs to be especially easy to use for those that don't dig social networking. There's a lot of us, and a lot of old school investors, just doing deals and not spending all of our time texting on our smart phones. Work on getting the best and the experienced here. A lot of them don't take kindly to the game thing, IMO.
I think putting "networking" before education/sharing limits who comes here and stays. I'm not here to network with maybe but a few people that are either in my areas of CA or doing something that complements what I do. I'm here because I like to share and learn new ideas. I've done some really complicated real estate deals and have a niche in probates and legal issues. I've even gotten title to props via adverse possession. I want to share what I know and learn something that takes it to the next level.
IMO, the site's lacking in active, experienced experts in several areas of deal making. Note investing and brokering, for one, which is a huge world of investors. I'd worry less about the Influence system and look at why some of those people aren't here. I'm pretty sure it doesn't have anything to do with the Influence system.
Very valid points K. Marie Poe
Steve Babiak - Good point.
Not sure where you want to end here Josh but I believe the consensus is - get rid of it.
Thanks for the feedback everyone. I'm going to go through the posts and will make a decision on how best to proceed.
I like the tracking of the number of the posts and I think the awards are fun. If someone "games" the system to get awards, that doesn't bother me, but its a little sad for them. As far as the possibility of folks misrepresenting themselves as a guru or reputable by "gaming" influence points; I think most internet users are savvy enough to take such ratings with a grain of salt. If it were my site, I would have a points system and awards, but I would not disclose the method. Rather I might just post a disclaimer for those who are that curious to just take it with a grain of salt...
I think it would be fun to have a ticker showing total votes a person has received tacked on with the number of posts ticker. Who cares if some folks get popularity votes? The reason you would be popular on this site is ultimately because you give valid, well thought out answers. For me, votes are the highest form of acknowledgement and affirmation to be had here on this forum, short of a written compliment. Then there is the issue of votes coming just because someone agrees with a post. I say that's a good reason to vote, too. Sometimes things get a little contested in here and its helpful to see what people may agree (or disagree) with by casting their vote, rather than a cumbersome written response saying "I agree" or a regurgitation of what is agreed with. So, I would be fine with a number of posts and a number of votes tally.
As an aside, I actually can't believe how much I have used this site, how much it has helped and also how personal it has become. I wish I could come to the summit to meet some of the people from this forum that have influenced decisions I have made in my "real" life. Maybe next year.
I think it would be fun to have a ticker showing total votes a person has received tacked on with the number of posts ticker.
I agree.
Ok - new system is in place . . . this is a 0-100 system and we've adjusted for your current influence. The algorithm will be refined over the next 2 weeks as we watch things come together. We think that the new scaling makes much more sense, and we have adjusted for time decay.
FYI - Everyone with zero influence now starts at 50 and negative influence will bring you to below the 50 . . .
We'll talk more about it soon -- time to drop some stuff at the hotel for the Summit!
I see one really obvious problem.
The user who started the thread I am linking to here had worse than -900 influence under the old system, and yet the reset puts her the same as any other member who has never violated the BP terms. Seems not quite right to me. High negative numbers should have been set to be below the mid-point of the new system IMO; we can debate the meaning of high in this context maybe.
http://www.biggerpockets.com/forums/97/topics/36164-protocol-for-purchase-of-reo-s-bank-notes
Nice! I think I may have found a bug already though :( I noticed on Jon K. 's post, it says follow him. I thought I already was. When I click on the follow Jon link, it takes me to the 'ooops, something went wrong page'
I agree though, kudos to you for getting this week with sooo many other important things on the table.
The user who started the thread I am linking to here had worse than -900 influence under the old system, and yet the reset puts her the same as any other member who has never violated the BP terms. Seems not quite right to me. High negative numbers should have been set to be below the mid-point of the new system IMO; we can debate the meaning of high in this context maybe.
http://www.biggerpockets.com/forums/97/topics/36164-protocol-for-purchase-of-reo-s-bank-notes
Perhaps that's because years have passed since the user referenced committed her offenses? Time served?