Welcome to NL Ice Data!

To complement the current Swiss hockey statistics resources, I decided to launch a new Swiss hockey website entirely focusing on hockey statistics and called NL Ice Data. The goal is to present statistics for the the National League, Swiss League, Junioren Elite A and Novizen Elite. By using most of the different statistics the SIHF displays on its website this platform aims at providing new interactive tools to improve the evaluation of teams, players, goalies and coaches' performances and to better understand the game of hockey:

  • Interactive tables and charts;
  • New available statistics and metrics for Swiss hockey;
  • Shot tracker maps;
  • A disciplinary database;
  • A glossary to explain the different terms used on this website and in the hockey analytics community.

    • New tools, projects and statistics along with some projects will be added to the platform. News and website's updates will be shared on the following Twitter accounts: @NLIceData | @spz19
    If the League publishes its proprietary data, in whatever format (.json, .xml or else), this website will be able to add some more *advanced* statistics for individual players.

    This website is the result of a few hundred hours of work as part of a project made under the supervision of Hockey-Graphs through their Mentorship program. Hockey-Graphs is one of the most renowned, and deservedly so, analytical hockey resources in North America and regularly publishing research about various hockey topics (contracts, tactics and much more).

    If you plan to use the different charts, table or data coming from this website, please link back to the website or Twitter accounts.
    There is a possibility that some developments/data/charts contents will be made available through subscriptions in the future. The subscriptions will be affordable and will be reinvested in the website.

    For any commercial use of the available content on this website, please send me an e-mail first at: nl.ice.data@gmail.com

    Do not hesitate to reach out would you have any remarks/suggestions or spot any error:

  • On Twitter: @NLIceData || @spz19
  • By e-mail: nl.ice.data@gmail.com

    • For other hockey statistics on Swiss hockey, you surely know the following classic resources:

  • www.sihf.com
  • www.planetehockey.com
  • www.hockeyfans.ch
  • www.eliteprospects.com
  • www.hockeydb.com
  • www.leading-sport.com

    • For some more *advanced* hockey statistics on Swiss hockey, you can also use the following resources:

  • Micha Hofer's hockey charts and viz with Tableau || Blog: Micha Hofer's blog on swisshockeyblogs.ch about advanced stats in Switzerland || Twitter: @EuroCaps
  • Daniel Weinberger's website about German and Swiss hockey || Twitter: @DanielT_W
  • Paul Yerly's blogspot (National League Stats) with interactive tables and charts || Twitter: @NLAStats
  • MySports' videos and articles on Advanced stats || Twitter (German) : @MySports_CH || Twitter (French) : @MySports_CH_fr
  • RSI Sports proposes some analyses with Andreas Hanni

    • Thank you for your visit!

    Players - Statistics











    to

    Players - Statistics - Game Logs








    Goaltenders - Statistics










    to

    Goaltenders - Statistics - Game Logs







    Teams - Statistics









    to

    * This table only reacts to the League, Season, Season | Series, Venue and Date filters *

    Players - Shot Tracker - Maps















    Explanations and use

  • The above graphs don't contain any shot attempt or goal directed on an empty net. As empty-netter shots or goals are often taken from the defensive or neutral zones, they could bias the numbers given in the above table and are only indicated as ENG in the last column. Here, the interest comes from shots in usual offensive situations.
  • For the moment, heatmaps color levels indicate the difference in shot distributions between players. A darker region as compared to another player may simply indicate that this player takes most of his shot from this region, not that he generates more shots from that region as compared to the other player.
  • The above maps only contain shots generated in the offensive zone.


    • Limitations

    Even if shot maps provide a substantial amount of information, it also misses important context such as:

  • Was the shot the optimal one at the moment of the shot?
  • Where were the opposite players?
  • Was the shot a screened shot?
  • Did a pass precede the shot?

    • As often in public sports websites, this missing data is not presented here because it is either not tracked or kept in-house by the leagues and clubs. Yet, the available data here is already a very good starting point for further analyses.

    Explanations and use

  • As for heatmaps, the above graphs above graphs don't contain any shot attempt or goal directed at the empty net.
  • Contrarily to the heatmaps, the graph is only reactive to the Player | Season | Season | Series and Strength fields.
  • To limit the presence of outliers, only players with more than 30 minutes of TOI are considered.
  • On the vertical axis, you can read ixG/shot, indicating the average expected goal value of a player's shots.
  • On the horizontal axis, you can read iCF/60, indicating the average number of shot attempts a player takes per 60 minutes of play. By comparing players in statistics per sixty minutes of play, you can compare their performances on an equivalent time basis.
  • In general, you'll find defensemen at the bottom of the graph, as shots are mostly taken from the blueline and carry less expected goal value than shots taken by forwards.
  • On the right, you'll find players that generate a high-volume of shots that carry on average less value.


    • Limitations

    Take for example Hofmann. He takes a lot of shot attempts and the average expected goal value is lower than most his forwards peers. Yet, it doesn't necessarily mean that in absolute numbers he takes less dangerous shots than them.

    Players - Shot Tracker - Statistics










    Goaltenders - Shot Tracker - Maps

















    Explanations

  • EQ OT (RS): by selecting *No* in this field, you can filter overtime situations at even-strength for Regular Season games (indeed, in the playoffs, the game at even-strength is still at played 5 on 5). As there is more space for players, as teams have already secured one point, the play is more aggressive and shots are, on average, more dangerous.
  • Strength: by selecting *Even-strength (+ extra attacker against)* you can filter even-strength situations where the opposite team pulled its goaltender and is trying to tie the game at the end of the game. There are important score effects in these situations and they could be considered as power-play situations.

    • These filters were added to better break down goalies' performances.

    Goaltenders - Shot Tracker - Statistics










    Explanations

  • EQ OT incl. (Reg. Season): by selecting *No* in this field, you can filter overtime situations at even-strength for Regular Season games (indeed, in the playoffs, the game at even-strength is still at played 5 on 5). As there is more space for players, as teams have already secured one point, the play is more aggressive and shots are, on average, more dangerous.
  • Strength: by selecting *Even-strength (+ extra attacker against)* you can filter even-strength situations where the opposite team pulled its goaltender and is trying to tie the game at the end of the game. There are important score effects in these situations and they could be considered as power-play situations.

    • These filters were added to better break down goalies' performances.

    Teams - Shot Tracker - Maps









    to








    to

    Explanations and use

  • The above graphs don't contain shot attempts or goals directed on an empty net. As empty-netter shots or goals are often taken from the defensive or neutral zones, they could bias offensive numbers and are only indicated as ENG in the last column. Yet, these numbers are displayed if you either select *Even-strength (+ extra attacker for)* or *Even-strength (+ extra attacker against)*.
  • For the moment, heatmaps color levels indicate the difference in shot distributions between teams. A darker region as compared to another team may simply indicate that this team takes most of his shot from this region, not that he generates more shots from that region as compared to the other team.
  • There can be some differences between data in this tab and the tab with teams' overall statistics, especially for the different Corsi statistics. The reasons will be further explained on a data quality article. Shortly, some teams overcount blocked shots, meanwhile some others (like GSHC or SCLT) undercounts them.

    • Filters - Explanations

  • EQ OT (RS): by selecting *No* in this field, you can filter overtime situations at even-strength for Regular Season games (indeed, in the playoffs, the game at even-strength is still at played 5 on 5). As there is more space for players, as teams have already secured one point, the play is more aggressive and shots are, on average, more dangerous.
  • Strength: by selecting *Even-strength (+ extra attacker for)* or *Even-strength (+ extra attacker against)* you can filter even-strength situations where one of the teams pulled its goaltender and is trying to tie the game at the end of the game. There are important score effects in these situations and they could be considered as power-play situations.

    • These filters were added to better break down teams' performances in different situations.

    Teams - Shot Tracker - Statistics







    to



    Filters - Explanations

  • EQ OT incl. (Reg. Season): by selecting *No* in this field, you can filter overtime situations at even-strength for Regular Season games (indeed, in the playoffs, the game at even-strength is still at played 5 on 5). As there is more space for players, as teams have already secured one point, the play is more aggressive and shots are, on average, more dangerous.
  • Strength: by selecting *Even-strength (+ extra attacker for)* or *Even-strength (+ extra attacker against)* you can filter even-strength situations where one of the teams pulled its goaltender and is trying to tie the game at the end of the game. There are important score effects in these situations and they could be considered as power-play situations.

    • These filters were added to better break down teams' performances in different situations.

    Score & Venue Adjustments

    Score adjustment

    The game is played differently whether a team is leading, tied or trailing. Leading teams will, on average, play sound and safe hockey, with a more defensive mindset, while trailing teams will, on average, be more aggressive and offensive. As such, trailing teams are expected to generate more shots than leading teams. This is called score effects and it needs to be taken into account. A team should not be overly penalized for trying to protect a lead.

    As such, shots taken by trailing teams have a lower weight than shots taken by leading teams. After applying these adjustments, numbers are more comparable between teams and better reflect their performances. It was also shown that Score & Venue adjusted statistics were the most predictive of goals and wins.

    Venue adjustment

    Home teams are, on average, expected to generate more shots than visiting teams. As such, adjusting for this effect will reduce the weight of shots taken by home teams, while increasing the weight of shots taken by visiting teams.

    Litterature

    You can also learn more about the score-adjustment methodology here or here in these pieces from Micah Blake McCurdy. In his pieces, Micah computes his adjustment coefficients by in-game situations (5v5). For the National League, the coefficients were computed using all situations, as the overall sample size of the data is lesser.

    Expected (xG) Goals Model

    If the Corsi, Fenwick or shots statistics are proxies for puck possession, an expected goals model will inform about the quality of the chances or the probability that shots will end up as goals. For example, A shot from the blue line will not have the same chances to end up as a goal as a rush chance.

    The current model is a slight variation of the model that was previously used for the Shot Quality charts for players. It includes the following variables: distance, angle, rebound, rush, penalty shot, offwing, as well as the strength state. Every shot attempt directed at an empty net have a zero xG value.

    Except for two minor changes made to the model, this is the same as the one used by Daniel Weinberger for his National League predictions. Thanks to him for sharing it.

    Data quality

    It's still a topic, but it deserves its own and complete article. Note that at the current state of things, HCD numbers, and more specifically its xG numbers, may be underrated. The same can be said for HCD players' indiviual performances.

    There can be some differences between data in this tab and the tab with teams' overall statistics, especially for the different Corsi statistics. The reasons will be further explained in the data quality article. Shortly, some teams overcount blocked shots, meanwhile some others (like GSHC, HCL or SCLT) undercounts them.

    Players | Goaltenders - Fouls












    Teams - Fouls











    Players - Shootouts