Tomorrow, the NFL kicks off another season and with that everyone from armchair quarterbacks to TV hosts and the President will be making predictions for games. Some people will pick on gut intuition, some on stats, while others pick by color or cuter quarterbacks. You may have your own method. This year, I am going to continue my work from last year in predicting winners of the games using social media buzz and some formulas using Lithium Social Media Monitoring.
Last season was interesting and I finished with a 132-123 record, followed up with a 7-4 record in the postseason, including correctly picking the Conference and Super Bowl winners. Despite the mild success in 2011, I wanted to go back to the drawing board and tweak the scoring system a bit.
This year I will be looking at data for 14-days for each team (Editor’s note – for Week 1, we will be using 7-day data) because looking at seven days last year skewed mentions when the byes started; and I will be dropping neutral mentions in terms of the scoring for NPS – this number grossly inflated scores for teams with national fan bases (Dallas, Pittsburgh, NY Giants) or teams that were a hotbed for national story-lines surrounding individual players (Tim Tebow). Below are the updated metrics and points breakdown:
- Net Promoter Score [NPS] (Positive Mentions – Negative Mentions/Weekly Team Mentions) – 3 Points
- Social Influence Marketing Score [SIM](NPS*Conversation Share*1,000) – 3 Points
- Conversation Share (Total Team Mentions/Mentions for Weekly Matchup) – 7 Points
- Positive Conversation Share (Positive Team Mentions/Mentions for Weekly Matchup) – 7 Points
- Negative Conversation Share (Negative Team Mentions/Mentions for Weekly Matchup) – -7 Points
In the case of a tie, the team with the most overall positive mentions will be selected as the week’s winner.
Also new for 2012 is a weekly challenge between me and some friends from all corners of the Interwebs. I will donate $100 to a charity selected by the individual with the best weekly score. The lineup includes:
- Week 1 Mike Pace
- Week 2 Mandie Boardman
- Week 3 Jason Mollica
- Week 4 Dave Folkens
- Week 5 Gini Dietrich
- Week 6 Matt LaCasse
- Week 7 Dan Haas
- Week 8 Marc Girolimetti
- Week 9 Amy Esposito
- Week 10 Nick Gosselin
- Week 11 Rebecca Denison
- Week 12 Bob Cargill
- Week 13 Jesse Bekas
- Week 14 Amanda Oleson
- Week 15 Chris Carfi
- Week 16 Shelly Kramer
- Week 17 Doug Haslam
What do you think of this test? Can you pick better? If you think so, feel free to play along by giving your picks in each week’s picks starting tomorrow (for special Wednesday opener) and then each Thursday after that.
[…] with last year’s tradition of using social media to predict this year’s weekly winners (click here for full methodology) against a group of friends who will use their own methodology to predict the […]
[…] been selecting each week’s winners using buzz from social media plugged into some math formulas (click here for more info). While last week’s buzz shocked me with the Arizona win over New England and a 9-7 […]
[…] and dump it into some formulas to assign a score which decides the winners. You can read more about it here. Each week we pit those social media picks against a friend of mine who picks using their own […]
[…] Social Media Monitoring, plug them into some formulas then *BOOM* winners are picked (more details here). To add to the intrigue, I ask a friend from around the Internets for their picks. The weekly […]
[…] (congrats on the acquisition guys), plug it into some formulas and poof there are the winners (click here for more info). To spice things up for 2012, we are putting our predictions against someone that I […]
[…] mark following a lackluster 6-8 finish last year. For t those of you visiting for the first time, click here to see our prediction method sponsored by Lithium Social Media […]
[…] formulas and like magic, unicorns predict the winners of the games (for a more detailed breakdown, click here). Each week we pair these results up against a friend of mine from across the Internet. This […]
[…] the format sounds odd to you, or you’re new here (click here for more), we pick the weekly games by pulling social media statistics and dump them into some […]
[…] media data pulled from Lithium Social Media Monitoring and dropping them into a series of formulas (full formulas here). During the week, we put our predictions up against a friend of mine in social […]
[…] Predicting Week 11 of the NFL Season Written on November 15, 2012 by Jeff Esposito in 2012 NFL Picks via Social Media, social media Hi there, thanks for stopping by. If you like this post, you might want to sign up to have my posts delivered to your RSS reader or sent to your email.Powered by WP Greet Box WordPress Plugin So far our social media NFL predictions have done OK, not great, but could be a playoff contender in the weak AFC this year. Currently we 77-68-1 (who ties games? It is like kissing your sister) after a 6-7-1 mark last week. For those of you coming to this blog for the first time, each week we pull 14 days of social media data from Lithium Social Media Monitoring and dump it into some formulas (more on those here). […]
[…] NFL season we predict the weekly outcomes using data from social media along with some formulas (more on that here). While this might sound odd, our social NFL predictions are doing decent 11 weeks into the season […]
[…] the post season. Over the course of the season we have been picking games using social media buzz (full data method here) and matched it up against a friend in the social web who picks using their own methodology. Some […]
[…] course of the season, we have been using social media to predict the outcomes of the weekly games (formulas here). On the season this method of predicting games has notched a 106-78-1 […]
[…] Monitoring and some formulas to predict the winner of each of the weekend’s games (more on that here). It is not scientific, but is more fun for a dork like me than say looking at jersey colors or […]
[…] With 15 weeks done in the NFL season, I am just happy that my Ravens backed into a playoff spot and I am hoping that they can win another game before the end of the season. In terms of our Social Media NFL predictions, we had a bad week 15 at 7-9, but have a decent season mark of 124-92-1. For those of you that are new here, we get dorky picking the NFL games and use social media and some formulas to pick winners (more here). […]
[…] the season we’ve used social media buzz to predict the outcome of each NFL game (full breakdown here). Despite a lackluster 7-9 showing in Week 16, this method has posted a 131-101-1 […]
[…] using social media buzz collected using Lithium Social Media Monitoring and a series of formulas (click here for full methodology). While this methodology might sound harebrained, it wrapped up the regular […]
[…] Lithium Social Media monitoring and dumped it into some weighted formulas to pop out the winners (more on the formulas here). While it is not science it helps mix two of my passions and hasn’t performed awful. On the […]
[…] to dump social data on the teams over a two week period and dump it into a slew of formulas (more here) and like magic out pop the winners. During the regular season, we wound up with a record of […]
[…] media buzz make the selection for me as I have throughout the regular season. The full method can be found here, over the course of the regular season; we racked up a 137-111-1 mark, but have sucked it up in the […]