Hacker News new | past | comments | ask | show | jobs | submit login

I suspect maybe you're showing (X exits)/(N funded companies). You can't just compare binomial proportions like that, between proportions of different N. You'll have lots of random fluctuation. If Waltham had a couple of exits out of few tries, it will look better than Cambridge, even though the error bars on Cambridge would be much smaller. Maybe do a lower bound of a confidence interval? Sorry for the stats nitpicks.



It was logistic regression, with the cities as features and exits vs. failures/floundering (1 vs. 0) as labels.


Wait, what? Logistic reg?

Can you explain that a bit, I don't see any classification going on

Anyway: I think cschmidt's point stands from what I can tell about your methods


It doesn't, as you can see from looking at the exit vs. total proportion for each place.


Yeah, while you can solve it that way, you're not getting any insight from the analysis. Logistic regression assumes all data are known exactly, while error bars are important for this case. Take a look at my comment to the parent for a better way.


Using the "error bar" approach (Wilson score), the following ranking results (need at least 7 startups to make the list):

Good:

        stddev          Exits   Total   Place

        3.17127075776   7       7       Chapel Hill, NC USA
        3.01843089923   143     207     Mountain View, CA USA

        2.94392395713   32      42      South San Francisco, CA USA
        2.33101381278   15      20      Foster City, CA USA
        2.31494411999   7       8       Itasca, IL USA
        2.31494411999   7       8       Westford, MA USA
        2.29564449793   539     966     San Francisco, CA USA
        2.2599048739    102     171     Cambridge, MA USA
        2.16920869781   37      58      Cupertino, CA USA
        2.15675433868   60      99      Waltham, MA USA
        2.14760333093   107     185     Santa Clara, CA USA
        2.03695302614   13      18      San Bruno, CA USA
        2.02879100247   8       10      Redwood Shores, CA USA

        1.96889748562   50      85      Menlo Park, CA USA
        1.92488214848   87      157     San Mateo, CA USA
        1.88561243541   111     206     San Jose, CA USA
        1.87088199283   12      17      Bedford, MA USA
        1.863724487     9       12      Brisbane, CA USA
        1.863724487     9       12      Alameda, CA USA
        1.81215794651   31      52      Burlington, MA USA
        1.80795794099   13      19      Los Gatos, CA USA
        1.79841328745   398     807     New York, NY USA
        1.79232968296   127     244     Palo Alto, CA USA
        1.78670614594   86      161     Boston, MA USA
        1.76603102228   130     252     Seattle, WA USA
        1.76116206965   14      21      Arlington, VA USA
        1.725668452     15      23      Aliso Viejo, CA USA
        1.69835027929   16      25      Emeryville, CA USA
        1.65992493077   115     228     San Diego, CA USA
        1.6288650578    21      35      Milpitas, CA USA
        1.62697011397   106     211     Sunnyvale, CA USA
        1.61878237122   8       11      Chelmsford, MA USA
        1.54045615814   9       13      Watertown, MA USA
        1.54045615814   9       13      Lowell, MA USA
        1.52765254628   23      40      Campbell, CA USA
        1.52184678546   63      124     Redwood City, CA USA
        1.45979095374   114     240     Austin, TX USA
        1.44275765525   12      19      Burlingame, CA USA
        1.43452996236   6       8       Calabasas, CA USA
        1.42071168128   15      25      Berlin, 16 DEU
        1.34271490531   7       10      Belmont, CA USA
        1.34271490531   7       10      Venice, CA USA
        1.34271490531   7       10      Louisville, CO USA
        1.33560665511   20      36      Pasadena, CA USA
        1.31683773572   18      32      Lexington, MA USA
        1.3079978015    35      69      Portland, OR USA
        1.29748756451   8       12      Sterling, VA USA
        1.28502636002   37      74      Tel Aviv, 5 ISR
        1.27560983396   9       14      El Segundo, CA USA
        1.26780216551   12      20      Bothell, WA USA
        1.26575403151   39      79      Fremont, CA USA
        1.25989899302   42      86      Santa Monica, CA USA
        1.16897845975   15      27      Marlborough, MA USA
        1.15041829453   30      61      Bellevue, WA USA
        1.14641742472   19      36      Morrisville, NC USA
        1.12663799508   18      34      Pleasanton, CA USA
        1.12663799508   18      34      Woburn, MA USA
        1.11830928242   39      83      Boulder, CO USA
        1.1055382851    17      32      Bethesda, MD USA
        1.1055382851    17      32      Richardson, TX USA
        1.0729130146    10      17      Malvern, PA USA
        1.06468478349   5       7       Kitchener, ON CAN
        1.06468478349   5       7       Surry Hills, 2 AUS
        1.06468478349   5       7       Bridgewater, NJ USA
        1.04813789306   27      56      Durham, NC USA
        1.0299173951    6       9       Petah Tiqva, 2 ISR
        1.02697161367   7       11      Tucson, AZ USA
        1.02697161367   7       11      Kfar Saba, 2 ISR
        1.00242961826   44      99      Vancouver, BC CAN
Bad:

        stddev          Exits   Total   Place

       -3.03488442406   13      135     Moscow, 48 RUS
       
       -2.24461906751   0       10      Saint Petersburg, 66 RUS
       -2.24054924343   2       22      Edinburgh, U8 GBR
       -2.1019848289    0       9       Porto Alegre, 23 BRA
       -2.1019848289    0       9       Little Rock, AR USA
       -2.1019848289    0       9       Jakarta, 4 IDN
       
       -1.99941240406   1       14      Lexington, KY USA
       -1.93525993713   0       8       Tallinn, 1 EST
       -1.93525993713   0       8       Centennial, CO USA
       -1.93525993713   0       8       Reno, NV USA
       -1.86110616417   5       30      Columbus, OH USA
       -1.74075145471   1       12      Glasgow, V2 GBR
       -1.73777813442   0       7       São Paulo, 27 BRA
       -1.73777813442   0       7       Taipei, 3 TWN
       -1.73777813442   0       7       Livermore, CA USA
       -1.73777813442   0       7       Brisbane, 4 AUS
       -1.73777813442   0       7       Turku, 15 FIN
       -1.73777813442   0       7       Lima, 15 PER
       -1.69885420792   2       16      Quebec, QC CAN
       -1.69885420792   2       16      Memphis, TN USA
       -1.65383249608   9       41      Cleveland, OH USA
       -1.62527946863   4       23      New Delhi, 7 IND
       -1.61418876295   6       30      Buenos Aires, 7 ARG
       -1.56634866608   55      178     Los Angeles, CA USA
       -1.53427233432   4       22      Chennai, 25 IND
       -1.48872091756   9       38      São Paulo, 2 BRA
       -1.40898404732   1       10      Colorado Springs, CO USA
       -1.40299548518   21      72      Dublin, 7 IRL
       -1.33066279643   4       20      Melbourne, 7 AUS
       -1.32829693893   130     356     London, H9 GBR
       -1.30824843556   26      83      Pittsburgh, PA USA
       -1.2681757003    13      46      Mumbai, 16 IND
       -1.24409522273   56      159     Paris, A8 FRA
       -1.21002433161   6       25      Hong Kong,  HKG
       -1.20538559097   1       9       Troy, MI USA
       -1.20094804147   12      42      Raleigh, NC USA
       -1.18989409055   24      74      Shanghai, 23 CHN
       -1.18887014907   45      128     Beijing, 22 CHN
       -1.11938393751   2       12      Fayetteville, AR USA
       -1.11331820129   19      59      Bangalore, 19 IND
       -1.10037104532   5       21      Phoenix, AZ USA
       -1.09606612444   35      99      Houston, TX USA
       -1.09541919563   29      84      Dallas, TX USA
       -1.09414918118   3       15      Santa Ana, CA USA
       -1.09414918118   3       15      Hyderabad, 2 IND
       -1.05291108276   16      50      Stockholm, 26 SWE
       -1.04636163325   59      155     Toronto, ON CAN
       -1.0386757642    13      42      Miami, FL USA
       -1.02924375409   10      34      Shenzhen, 30 CHN
       -1.00129935812   9       31      Calgary, AB CAN


Moscow, "a success city"... =D




Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: