# More Statistical Fun (OR NOT) Top Trulia Voices profile page lists top ten for 3 categories:

Asked by Jim Walker, Carmichael, CA Tue Aug 28, 2007

Top Trulia Voices profile page lists top ten for 3 categories:

Most Liked

I did a statistical analysis to show another way at looking at Trulia production.
I leave it to the crowd to draw their own conclusions and comments. Please be kind to each other. My intent is not to laud or disparage anyone. I am just crunching the numbers that Trulia already has provided us in a different formula, which is:
(U plus B divided by M= Score) translated:
Useful Thumbs plus Best Answers divided by total # of answers equals Score.
Useful, Best and Most Questions are not addressed by this analysis.
I am including only ( top ten) posters who have answered more than 200 questions.

Kaye 299/260 = 1.15 | Deborah 648/580 = 1.12 |
Bridgette 278/294 = .94 | Jeanette 320/346 = .92 |
Keith 163/206 = .79 | Carrie 226/332 = .68 |
Trulia Roger 141/224 = .64 | Cindi 113/277 = .40
Bruce 158/463 = ..34 | Pam 153/543 = .28

## Help the community by answering this question:

16
I have an analytical orientation, so I certainly have been cognizant of the relationship between the most answers , TU, and BA counts. I had never formally calculated any ratios for a list any Trulians. I donâ€™t answer or not answer a question predicated upon my expectation of how it will affect my individual count, or ratio.

If my ratios included result numbers that cautioned me to take more notice in this area, I would. So, I am not claiming that I turn a deaf ear. If the ratios sounded an alarm, I would more carefully think about each question before I answered. I just â€œdo what I doâ€ and will continue along that path for the while. At times, that includes avoiding involvement with certain threads.

Thanks, Jim, for taking the time to put things in perspective. While none of the ratios came as a big surprise, it was interesting and provided clarification and validation.

Ruth, Tx for the compliment. And, Ruth, you make an important point about the fact that some questions may go unanswered without a first answer acknowledgment. This will remain important in expanding public awareness. I have seen Trulia use speed of answer (20 minutes) in promoting the value of Voices to the public.
1 vote Thank Flag Link Wed Aug 29, 2007
MVP'08
Contact
I think Trulia should take your statistics into consideration. I have noticed that there are many agents who either repeat something previously stated by another (against community guidelines) or don't really answer the question (another no-no) just to get their numbers up (I'm assuming. Otherwise, why bother being redundant and ineffective?)
It may be that it's so important to be a Trulia Top Voice, but even that rating system seems off to me. When I search my zip code (the one I registered under), I am the only agent there. Yet when one searches Top Voices for my zip code, other agents names pop up who are located no where near me.???
Anyway, bottom line is that I think that context and content (quality) are much more important than quantity and there should be a way to recognize the agents who deliver the good answers over those that just repeat someone elses answers! I love your system!
Web Reference: http://www.dianeglander.com
I'll have to delve into that further but want to point out the skew factor. When Trulia started people were using it differently or there weren't enough participants to make the analysis fair. Looking at "last month" is much different than the entire picture. This was obvious to me when I looked at the number of Best Answers Laurie received and she hasn't posted since July.

I also discovered (I think) that posting follow up comments to the same question doesn't increase your count. But the number of thumbs up can exceed your number of posts.
Ruth
MVP'08
Jim..
This is certainly interesting and something I would never have thought about.. I answer broad questions or those about my area. I never answer a specific question about a specific area. I may know about making an offer but not about the best schools in KS.

Most liked answers can be influenced by "friends" but not the best answer. One thing I have noticed since Trulia got larger is that the person who asks a question often doesn't note which answer he/she considers to be the best answer.. this is something only the poster can do so it may explain why there are not as many best answers as there should be.. especially considering the number of answers many posts receive.

That said... I like my bear and I'm not giving it back.... LOL
MVP'08
Contact
Wow. What a bunch of party animals. Somebody stop them before things get out of control ;o)
I forgot to metion Jim's score 1.30 - I am going to stop now, it's diluting my score.
MVP'08
Contact
Thanks Sylvia

I know you conducted your analysis with top 10 voice over 200 answers.

The number of answers is key. at least more key than I considered earlier.

Lets say a new Realtor joins Trulia. We will call them MUROFOBSF. ( I cant resist )

MUROFOBSF just happens to have three friends who give three thumbs up.
(u) 3 + (B) 0 / (m) 1 = 3

Again I know we are in the top ten and over 200 answers.

However I have to amend my previous answer.

I now feel that this system could be manipulated as much or more than the current system

As far as Sylvia`s Idea to give more weight to the best answer, She has a point.

How many points I don`t know?
MVP'08
I don`t know what to say Ruth. You have had a very busy day in Trulialand.
I truly do appreciate the Detective work you have done. Hopefully Trulia has been reading.
You even had time today to tell me how to change my photo, which it still dosn`t work.

And now on to the Numbers.

I have spot checked your math. ( Real Estate Master IIIx )
I have read a lot of the answers by these posters.( over the last two months)
My conclusion

Your statistical analysis is in theory much more fair, and correct than the current system.
I am not saying that it can not be manipulated, just not as much as the current system.

You are pretty smart for a painting.
MVP'08
Jim, Great statistics. I have looked at that also. I would have done something very similiar and did for just myself once. I noticed that my percentage of quality could be a bit better. I also think that as we become more acquainted with the site and it participants, we adjust our responses. I have discovered the same thing as you Ruth on the multiple answers to questions. Thanks for you input.
Web Reference: http://carriecrowell.com
Well, I'm only about half way to being considered and rated by Jim's fun statistical twist.

But this makes me realize that I never even consider the points when answering because my aim is to give quality answers that will genuinely benefit the buyers and sellers and those considering entering the market. One benefit of this is that I've had my answers re-posted on websites here and there (which I discovered by Googling), and that reinforces my belief that it is the CONTENT THAT COUNTS!

But then, if I got a high score from Jim, I'd be ticked by that too:)
1 vote Thank Flag Link Wed Aug 29, 2007
Now that i am reading Deborah's question, I hope you guys/gals did not misunderstand what I said in my previous answer, which I think could have been misunderstood.

________________ I said___________________

Just what I strive for, not always successful, sadly: Best answers, Useful answers on the questions I am interested in and I think I can contribute to - again, if there are tons of great answers; which are many; and I won't make much differences; then I won't answer that question!
_______________________________

What I meant was that when I answer any question, I want to answer it in the way that are useful and meaningful; and hopefully is an answer which best answers to what the person is asking.

I did not mean to say that I try to answer the quesiton in a way so that my answer is a popular answer to get most votes. That is never my goal. I hope that's not the impression I am giving you by what I said.

I also think that's part of reason why Jim posted the statistics - to draw the attention between quality and quantity.

Sylvia .
1 vote Thank Flag Link Wed Aug 29, 2007
MVP'08
Contact
I think Most Answers is useful to discredit people who talk without a purpose. When someone answers more questions than anyone else and doesn't even rank in the top ten for Useful Answers, you loose respect for them. Then there is the flip side, Paul doesn't even rank in the top ten for most answers yet he'll have top Best Answers. Deborah is reliable. If you need an answer, you can count on Deborah answering and getting excellent advice.

I thought First Answer should be eliminated until I got my Bear from Trulia. Their note said on average questions are answered within 20 minutes. I don't think some people would get any answers if Trulia didn't reward first to answer.
1 vote Thank Flag Link Wed Aug 29, 2007
MVP'08
Hi Jim:

Great analysis. Just what I strive for, not always successful, sadly: Best answers, Useful answers on the questions I am interested in and I think I can contribute to - again, if there are tons of great answers; which are many; and I won't make much differences; then I won't answer that question!

However, keep in mind that everything is subjective and a lot of times really depend on when the question is posted, who answered in what order and whom the question draws attention to, which all make a difference in how it was voted. Some great answers donâ€™t get as much attention as witty answers or what not. However, in the long run, by law of largeness, it shall all evens out.

I do think you might want to consider giving more weight to the Best Answer. An answer can get tons of thumbs up because of it is either really useful or because some of the abovev reasons, but the Best Answer, hopefully, is the one that either answers or closet to answer the originatorâ€™s question. So, I think Best Answer should weight more.

Then again, did you look at Paulâ€™s statistics? Super knowledgeable, great agent from what I can see; seldom wastes his time on this kind of questions. His number is 1.2849, Best Answers not weighted.

Sylvia
1 vote Thank Flag Link Tue Aug 28, 2007
MVP'08
Contact
I am feeling a bit lame at the moment! (But Jim, I ike your thoughts...)
Web Reference: http://www.cindihagley.com
1 vote Thank Flag Link Tue Aug 28, 2007
MVP'08
Contact
Diane, I agree (posting for ratings) but I have been corrected twice now. I gave one person a thumbs down for doing that and told her in the post. Her response was that she thought the questions was a poll type question and she was right. The person asking the question was looking for a consensus.

Then another time I started to point it out to someone who annoys me. I looked through that person's posts from his/her profile and couldn't find a glaring example or habitual pattern. I got a different feel just reading one person's posts verses reading that person's post amongst everyone else.

Many times I have to hold back from responding because what I say will be redundant. It kills me that I just read 12 responses to a question (I'm an extremely slow reader) and can't even give my stamp of approval. That's why I was pleased to discover that the back and forth comments within a question, only gets counted as one answer.

Oh, wait, another reason I've been wrong. When you write as slow as I do, or when I do research for a question, by the time I post two people have already said the same thing.

The zip code search is something I don't think Trulia expected to happen. Trulia's mission statement (or something like that so I'm told) is that real estate is local. However, the value of the forum is learning how other states and regions handle things. When names pop up that aren't in the same state, it's because they answered questions from people who put in their zip code.

Ruth
1 vote Thank Flag Link Tue Aug 28, 2007
MVP'08
I wrote the following addendum this morning before anyone else had responded so the numbers may have changed ever so slightly, but they are close enough.

As I stated in the opening of this thread â€œMy intent is not to laud or disparage anyone. I am just crunching the numbers that Trulia already has provided us in a different formulaâ€

The consensus from earlier threads that the â€œusefulâ€ â€œnot usefulâ€ and â€œbest answersâ€ are not scientific or objective in any significant way. However, when numbers are lying around on the floor they are passively asking to be analyzed, just as a sleepy golden retriever passively asks to petted just by lying at your feet.

I only analyzed the top ten most prolific answerers.
The study is flawed in that I didnâ€™t further analyze best most and most liked questions or â€œfirst answersâ€

I might recommend Trulia drop the first answer category as it seems to encourage speed of posting over careful construction of a complete and useful answer.

Another flaw in the study may be that posters in the earliest days of Trulia picked up a very high ratio of Best Answer and Useful Answer Thumbs per answer. I did not analyze this. It is a casual observation. I suspect that there are software tools more sophisticated than my \$2 Casio calculator that can give a month by month analysis.

There was no room in the original post; here are some of the posters who have placed between 100 and 200 answers and their U plus B divided by M ratio, I name the ratio UTBA/TA Score

Brian Brady 348 / 134 = 2.60 ||
Mitchell Hall 304 / 130 = 2.34 ||
Roberta Murphy 340/181 = 1.88 ||
Maureen Francis 306/ 169 = 1.81 ||
Paul Slaybaugh 216 / 177 = 1.22 ||
Sylvia Barry 213 / 196 = 1.09 ||
Ruth J 161/152 = 1.05 ||
Ellen Chung 148 / 144 = 1.02 ||
Ted Shoop 75 / 140 = .54 ||
Herman Chung 13 / 154 = .08 ||

The study may also skew the Utba ratio higher for posters who answer general questions that have a lot of readers, especially if one posts well early. Some of the good answers posted early have gotten as many as ten thumbs up and a best answer from a highly attended thread.

A major goal of Trulia seems to be to answer consumers specific questions about a metro area, those sort of questions may not be well attended by out of area readers and thus get fewer positive thumbs than a general question. Yet, these are extremely important and a good answer about a zip code or neighborhood is worth a lot more to the Trulia consumers than some funny joke I might tell. Yet my joke might score higher.

PS. Patrick, in the morning I am going to check MUROFOBSF's UTBA-TA Score!
MVP'08
Contact

View all »