Previous studies have shown that RI is necessary to measure the possibility of tourists’ revisit. Now, as a prediction tool for tourists’ revisit, it has become a hot topic in tourism circles (Assaker et al., 2011), and scholars have paid attention to the study of tourists’ RI. For example, He and Luo (2020) built a model of travel motivation, SAT and RI based on the push-pull theory and verified the relationship between tourists’ travel experience and RI. Soliman (2021) expanded the TPB model and constructed travel motivation, online word of mouth and destination familiarity to predict tourists’ RI. Tajeddini et al. (2021) extended the TPB model to models of SN (peer influence, word of mouth), ATT (social influence and economic influence) and quality (service quality and experience quality) for a prediction of tourists’ RI. Research has found that revisiting tourists provides a long-term and stable source of income for the tourist destination and a high proportion of tourists, which helps to maintain the tourist source share of the tourist destination. In terms of long-term cost, retaining fixed customers is far lower than acquiring new customers (Petrick, 2005). Moreover, a 5% increase in customer retention can increase profits by 85% (Reichheld, 1993). From the perspective of event operation cost, the development cost of the revisit market is lower than that of the initial tour market, and improving the revisit rate of event tourists is conducive to the sustainable development of sports event tourism, the economic and social development of the event’s host place (Chen et al., 2021).

According to the TPB model, SNs are another factor affecting behavioural willingness. Subjective norms refer to an individual’s perception of social normative pressures that compel them to enact a particular behaviour or the people important to him or her to engage in or refrain from a specific behaviour (Ajzen, 1991; Ajzen & Fishbein, 1975; Belanche et al., 2019; Tajeddini et al., 2021). Whether or not an action is performed is influenced mainly by the people who have a close relationship with the subject. Regarding tourism, certain research has demonstrated that the previously published literature empirically proves that travellers’ views influenced people’s behavioural intentions, that is, close people (family, friends or colleagues) influence their choices, and that SNs have a significant positive correlation with RI (Abbasi et al., 2021; Soliman, 2021; Tajeddini et al., 2021). This leads to the following hypothesis:

Perceived behavioural control is also considered a prerequisite for intention and behaviour (Ajzen, 1991), defined as the perception of how easy or difficult it is to act (Song et al., 2015). In other words, PBC means that the individual believes that they can perform the behaviour skillfully. Primarily, PBC considers individuals’ perceptions of how well they manage aspects that may enable or limit explicit behaviour. Perceived behaviour control also positively impacts individuals’ actual access behaviour (Soliman, 2021; Tajeddini et al., 2021). In addition, studies have shown that the growth of PBC will increase the endurance of revisiting the destination (Lam & Hsu, 2006). In addition, RI is positively and significantly correlated with PBC (Abbasi et al., 2021; Han et al., 2010, 2017). Based on the earlier discussion, the following hypothesis is formulated:

Attitude refers to the evaluation (positive or negative) of an individual’s willingness to engage in the behaviour of interest (Ajzen, 1991). According to Schiffman and Kanuk (2004), theoretically, ATT consists of three components: cognition, emotion, and intention or behaviour. The cognitive component relates to a person’s cognition, which is derived from direct experience or various information sources that form an ATT and are expressed as beliefs. The emotional component is a mental response that explains preferences, emotions and feelings. Cognition indicates the likelihood or improbability that a person will take a particular action or act in a particular way, indicating future intentions. Personal experiences, family, friends, the internet and the media may influence a person’s ATT. Over time, ATT form and remain consistent. It shows a tendency to act in a particular way as a response to the influence of past knowledge and experience (Lam & Hsu, 2006). Based on these beliefs, Reza Jalilvand et al. (2012) carried out projects to measure whether tourists’ ATT towards a destination are ‘very bad/very good’, ‘very worthless/very valuable’ and ‘very unpleasant/very pleasant’. It can be seen that ATT affects a person’s behavioural intention. Research has found a significant positive correlation in tourism between the ATT towards visiting a specific destination and the intention to see that destination (Abbasi et al., 2021; Han et al., 2010). Therefore, we propose the following hypothesis:

Regarding sports events, many scholars try to develop reliable and effective service quality measurement scales. Currently, SERVQUAL and SERVPERF are the two main representatives of the scale (Jain & Gupta, 2004). In sports and leisure, researchers have used a modified version of SERVQUAL to measure the quality of service over the past few decades. For example, McDonald et al. created TEAMQUAL with 39 items for sports organisations based on SERVQUAL (McDonald et al., 1995). On the other hand, Theodorakis et al. (2001) developed the SPORTSERV scale, which represents five service quality dimensions: tangibility, responsiveness, accessibility, security and reliability. Based on multiple focus group interviews and an extensive literature review, Ko et al. (2011) developed a comprehensive model and measurement SEQSS, which consists of five dimensions: game quality (GQ), augmented service quality, interaction quality (IQ), outcome quality (OQ) and environment quality. Jin et al. (2013) proposed a four-dimensional framework composed of GQ, IQ, OQ and physical environment quality (PEQ). The GQ refers to the audience’s evaluation of the players’ skills and information. Interaction quality involves evaluations of the staff and volunteers’ responsibility for the service and SAT with interactions with other audiences. Outcome quality provides a ‘measure of consumer perception of benefits gained from interactions with service providers’, while PEQ focusses on consumer evaluations of the design, cleanliness and safety features of the stadium atmosphere and facilities. Ko et al. (2023) developed 12 individual measurements based on the dimension of ETQ and found that the utility and effectiveness of the scale could be significantly improved. In addition, as there is little research on the quality of major sports events (Jin et al., 2013), in order to understand the psychology and behaviour of sports tourists, our research focusses on the quality of events rather than the quality of services.

Quality is considered as perception (after the service experience) minus expectation (before the service experience; Tajeddini et al., 2021). Previous studies have found that ETQ positively influences RI (Moon et al., 2013). Wicaksono et al. (2021) predicted tourists’ RI by using a measurement scale for the quality of sports events, which was composed of three dimensions. Jin et al. (2013) proposed that ETQ is a four-dimensional framework and found that sports events have a specific positive impact on tourists’ behavioural intentions. It can be seen that tourists feel that the ETQ is higher and superior, and they are likely to increase their chances of revisiting the same destination. Therefore, based on the earlier discussion, the following hypothesis is proposed:

A large number of literature focusses on the relationship between perceived quality (Vegara-Ferri et al., 2020), experience quality (Aliedan et al., 2021; Vegara-Ferri et al., 2020), service quality (Al-Laymoun et al., 2020; Elfi Azhar et al., 2019) and SAT of tourism. However, studies on ETQ are still lacking. In recent years, in the context of sports tourism, more and more studies have paid attention to ETQ. For example, Chen et al. (2023) studied the determinants of sports tourists’ decision-making and discovered that SAT and ETQ are related. In sports tourism, Jeong et al., (2019) and Jeong & Kim, (2019a) investigated the relationship between SAT, ETQ and behavioural intention and found that tourist SAT is positively correlated with ETQ. Lam-González et al. (2021) studied the relationship between high-quality urban entertainment, destination image, experience quality and tourists’ SAT and found that the tourists’ SAT and ETQ are correlated. Given the existing research on the quality and SAT of events, we propose the following hypothesis:

The SAT of people who purchase services is generated by the travel experience, which leads to SAT with the services received (Gillison & Reynolds, 2018). Satisfaction refers to the emotional response to a cognitive response to an experience (Monferrer et al., 2019; Smith, 2020). Studies in the tourism context have established a link between SAT and willingness to revisit, concluding that SAT significantly impacts tourists’ positive willingness to revisit (Abbasi et al., 2021). Jeong et al., (2019) examined the structural relationship between ETQ, tourism SAT, place attachment and behavioural intention, which predict behavioural intention. Hutchinson et al. (2009) developed and tested a comprehensive model to examine the relationship between golf travellers’ perceptions of quality, value, fairness and SAT, and the findings indicated that the intention to return was significantly influenced by SAT. Meanwhile, Tang et al. (2022) found that tourist SAT directly and positively impacts tourists’ RI through rural tourism. As a result, the study by Bayih and Singh (2020) showed that when tourists are dissatisfied with a destination or its activities, their willingness to return decreases significantly. To sum up, if tourists think that the quality of sports event service is higher and better, they may have higher SAT, which is likely to increase their chances of visiting again. Therefore, based on the previous discussion, we propose the following hypothesis:

In addition, according to the study on the relationship between ETQ and SAT, combined with the RI, Jeong et al. (2019) found tourist SAT’s moderating role in the association between behavioural intention and ETQ. Therefore, based on the previous discussion, the following hypothesis is proposed:

The extended TPB model shown in Figure 1 is constructed based on the aforementioned results.

In July 2022, Taipan Village, Taipan Township, Qiandongnan Prefecture, Guizhou province, held the first Guizhou ‘Beautiful Village’ Basketball League, which became popular at home and abroad through short videos in just 1 month. The local villages that held the sports event are affectionately called ‘Village BA’ by netizens. This is precisely because the ‘Village BA’ basketball tournament in Guizhou has the functions of sports, tourism, national culture and commercial consumption, which meets the current market demand and the psychological needs of sports tourists. From July 2022 to August 2023, Taijiang County in Guizhou received more than 2 million tourists and achieved a tourism income of more than 2.3 billion yuan. At present, the ‘Village BA’ basketball tournament in Guizhou can stand out among many sports events in China, become a viral sports event at home and abroad, gain sustainable competitive advantages, attract a large number of sports tourists to Taijiang to watch sports events and play a significant role in promoting local economic and social development. However, there are few empirical studies on the tourist behaviour of basketball events before, so we will take tourists who choose to visit the ‘Village BA’ basketball tournament in Guizhou as the research object to understand better the potential and unexpressed factors that affect tourists’ choice to travel to Guizhou, China.

To collect data, we conducted an exhaustive literature review to select the most appropriate scale for questionnaire design to measure different concepts. Finally, the questionnaire design includes the main body scale of tourists’ basic information and research structure. Among them, the basic information about tourists involves demographic characteristics, such as gender, age, occupation, income level, education level, etc. The topic scale refers to research scales widely used and validated in the existing literature. In combination with interviews with sports tourism practitioners, tourists and related scholars, these indicators have been slightly modified to suit this study, and the model consists of six constructs, including one second-order construct and five reflective formative first-order constructs (Table 1). To be specific, four items were used to evaluate the ETQ (Jeong & Kim, 2020, 2022); Jeong and Kim (2020) used item segmentation to transform subfactors of ETQ into four observable variables, for example, the GQ includes three items, the IQ consists of three items and the OQ contains three items. The PEQ consists of three items. Three items were used to evaluate the tourist SAT (Jeong & Kim, 2019b, 2020). Tourists’ ATT (three items), SN (three items) and perceived behaviour control (three items) were originated from researchers (Liu et al., 2020; Meng & Cui, 2020; Soliman, 2021; Wang et al., 2020). Revisit intention (three items) originated from Jeong et al. (2019). Before the prediction, a panel of three professors in sports tourism was invited to conduct a review to check the surface validity of the questionnaire items. Then, a 7-point Likert scale was used for all structures, with 1 being ‘strongly disagree’ and 7 being ‘strongly agree’. In addition to ATT and RI, this was measured using a 7-point semantic difference scale (Japutra et al., 2019; Quintal et al., 2015; Vesci & Botti, 2019).

In this article, SPSS 26.0 is used to calculate the data collected by the questionnaire. We tested these hypotheses using partial least squares structural equation modelling (PLS-SEM) of SmartPLS version 4.1.0. We chose to use SEM-PLS for three main reasons. Firstly, this study focussed on prediction by testing the variance of endogenous variables and identifying the critical predictors of the variables. Secondly, the measurement model includes one formatively measured construct. Thirdly, PLS-SEM was to understand theoretical extensions of established theory better. Lastly, the structural model can be easily be estimated with a higher-order construct (Hair et al., 2017). Partial least squares structural equation modelling consists of a two-step estimation process: measurement model assessment and structural model assessment. In the first stage, the measurement model involved evaluating indicators’ reliability and validity, such as outer loadings, internal consistency reliability, composite reliability (CR), convergent validity of each construct and discriminant validity. The structural model was checked for potential collinearity issues among the constructs in the second stage. Then, the structural model assessment included learning about the predictive capabilities of the model and determining the predictive power of hypothetical models, such as coefficient of determination (R²), cross-validated redundancy (Q²) and the path coefficients (Hair et al., 2013, 2019; Sarstedt et al., 2017).

The convenient sampling method has been widely recognised and used in the field of sports tourism (Alexandris et al., 2017; Chen et al., 2021; Jeong & Kim, 2020), because it offers several advantages, particularly in situations where time, resources or access to random samples are limited (Emerson, 2021; Etikan, 2016). Therefore, a convenient sampling method was adopted in this study. According to Krejcie and Morgan (1970), the sample size selection should be 385 for large population size. In addition, Nunkoo et al. (2013) and Hair et al. (2019) suggest that the minimum sample size in SEM is 200, and their views have been widely applied in empirical studies. Considering the questionnaire recovery rate and efficiency, and further ensuring the sufficient sample size, we finally chose 500 tourists to Guizhou sports tourism destination as the sample size. From 15–18 July 2023, a 4-day field survey was conducted on ‘Village BA’ in Taipan, involving more than 40 people, including cadres of the Taipan village committee, 9 tourists, referees and athletes, and they filled in relevant questionnaires. In addition, the objects of this study were mainly from all over Guizhou province. Because of the younger basketball population, a structured self-filling network questionnaire is used to collect data. The network questionnaire is efficient and cost-effective, and the respondents can truthfully express their opinions without being affected by other factors. This article selects the online survey platform ‘Wenjuanxing’ to publish and collect questionnaires. The validity date is from 01 October 2023 to 20 December 2023 and a link is sent to invite tourists who watched the ‘Village BA’ basketball match in Guizhou to fill in the online questionnaire. A total of 500 questionnaires were distributed. The researchers excluded the questionnaires with missing values and the questionnaires with similar answers, and the number of valid questionnaires was 367, with an effective rate of 73.4%. According to the criterion that the adequate sample size is 10 times higher than the number of indicators (Thomspon, 2000), the effective sample size of this time meets the standard for subsequent statistical analysis. The demographic characteristics of tourists are as follows: regarding gender distribution, the proportion of male tourists is 46.9%, and that of female tourists is 53.1%. In terms of age distribution, 32.4% were under the age of 25 years, 18.5% were between 26 and 30 years, 19.3% were between 31 and 40 years, 13.4% were between 41 and 50 years, and 16.3% were over 50 years. In the distribution of education level, high school education and below accounted for 12.6%, junior college accounted for 18.5%, university undergraduate accounted for 49.6% and postgraduate education accounted for 16.6%. In terms of occupational composition, 28.6% were full-time students, 21.5% were freelancers, 20.2% were self-employed persons and 7.6% were technical development and engineers. In addition, 1.4% were lawyers or legal affairs, 6.3% were teachers, 6.8% were enterprise managers, 2.5% were scientific researchers and 2.2% were government personnel; retirement accounts for 1.6%. In terms of monthly income, 65.7% of people have an income of less than RMB 5000 Yuan and 34.3% have an income of more than RMB 5000 Yuan. The overall questionnaire sample is representative to some extent.

Ethical clearance to conduct this study was obtained from the Guizhou Education University Human Research Ethics Committee.

This study used two stages to evaluate the measurement model. In the first stage, in order to verify whether the measurement scale of nine reflective exogenous constructs, including SN, PBC, ATT, GQ, IQ, OQ, PEQ, SAT, RI, were evaluated in line with their internal consistency reliability and validity. Therefore, we checked and reported the outloading of all items for each construct, Cronbach’s α and CR, and average variance extracted (AVE) (Hair et al., 2017). Table 1 shows the internal consistency, reliability and convergence validity results.

As seen from Table 2, the Cronbach’s α values of each variable were above the 0.70 recommended threshold (Hair et al., 2019) except SN factor (α = 0.693). However, the coefficient α values in the range of 0.6–0.7 could be considered as the minimum acceptable confidence level for preliminary studies (Kim et al., 2015). Therefore, factors of SN are retained in this study. The range of CR is between 0.752 and 0.897, which is greater than the common standard of 0.7 (Hair et al., 2019). It indicates that the reliability of the scale is high. The outloading of all items ranges from 0.702 to 0.918, all greater than 0.6 (Hair et al., 2017), and the AVE ranges between 0.619 and 0.817, greater than the standard of 0.5 (Hair et al., 2019). This indicates that the scale has good internal consistency and high convergent validity.

In order to investigate discriminant validity, we applied the approach called the Fornell–Larcker criterion (Fornell & Larcker, 1981) and heterotrait–monotrait (HTMT) ratio (Hair et al., 2019; Henseler et al., 2015). The discriminant validity was evaluated by comparing the AVE value and squared correlation with other constructs (Fornell & Larcker, 1981); many scholars used the Fornell–Larcker criterion to analyse the discriminant validity (Erul et al., 2020; Liu et al., 2021; Meng & Choi, 2019). Meanwhile, HTMT is the most popular measurement method at present. Heterotrait–monotrait was defined as the average of item correlations across structures relative to the average of items measuring the same structure (Henseler et al., 2015). Table 3 shows the results. The AVE values of all constructs are higher than the squared correlation with other constructs (Fornell & Larcker, 1981). Besides, the HTMT ratio values ranged between 0.187 and 0.663, which were below the threshold of HTMT_0.85 or HTMT_0.9, with the average of item correlations conforming to conceptually different structural HTMT_0.85 recommended value (Hair et al., 2019). Therefore, the scale has discriminant validity.

In the second stage, GQ, IQ, OQ and PEQ established ETQ as a second-order formative construct. These four first-order structures become observation indicators of ETQ. For the structure of formative measures, convergent validity was assessed by correlating the structure to an alternative measure of the same concept (Hair et al., 2019). Therefore, the first-order structures’ variance inflation factor (VIF) is used to test the collinearity of formative indicators. As can be seen from Table 4, all VIF values are between 1.43 and 1.973, below the threshold of 3.0 (Hair et al., 2019). In addition, Table 4 showed that the outer weights were all significant except for IQ (p > 0.05), and the loadings of four first-order constructs exceeded 0.50. According to the suggestion by Sarstedt et al. (2017), the weight as soon as possible was not significant, but its loading was greater than 0.5, so we retained the IQ structure. As a result, the formative-reflective construct was established, indicating that the measurement items adequately measured its respective second-order construct.

The structural model was evaluated to determine the significance of the hypothetical path and the overall predictive capabilities of the model, as indicated by the following criteria: coefficient of determination (R²), cross-validated redundancy (Q²) and the path coefficients (Sarstedt et al., 2017). In the first stage, a bootstrapping procedure was performed on 5000 samples to assess the importance of the hypothetical path. In the partial least squares (PLS) structural model, the coefficient of determination (R²) is referred to as the in-sample predictive power (Rigdon, 2012), and the R² values were 0.25, 0.50 and 0.75, respectively, which are considered as weak, moderate and substantial effects (Hair et al., 2017; Sarstedt et al., 2017). In this study, we used the corresponding exogenous variables to explain the R² value of the variance of endogenous variables, and the results show that tourists’ SAT explained 18.6%, and RI explained 56.4% of the variance in the model. Therefore, this result was considered acceptable (Rasoolimanesh et al., 2019). Furthermore, according to a recent study (Sarstedt et al., 2017), the model fit was assessed using the value of standardised root mean square residual (SRMR). The value of SRMR was 0.084. This was close to the required value of 0.08, indicating an acceptable model fit (Hu & Bentler, 1998; Wan et al., 2021).

In the second stage, 95% confidence intervals were used for hypotheses testing to evaluate the sign and significance of the path coefficients (Ali et al., 2018; Tajeddini et al., 2021). Table 5 shows the evaluation of the structural model. Constructs of the extended TPB were positively associated with RI for sports tourism. Attitude (H1: β = 0.279, p < 0.001, t = 5.421) had the strongest effect on RI, followed by SN (H6: β = 0.225, p < 0.001, t = 4.56), PBC (H4: β = 0.204, p < 0.001, t = 4.628), ETQ (H2: β = 0.181, p < 0.001, t = 3.685) and SAT (H5: β = 0.139, p < 0.01, t = 2.622). Thereby, the findings support H2, H3, H4, H5 and H6. In addition, ETQ (H3: β = 0.432, p < 0.001, t = 8.796) had the strongest effect on SAT. Thus, findings support H3. Then, mediation analysis was performed to assess the mediating role of SAT. The result (see Table 3) revealed a partially significant (p < 0.05) mediating role of SAT (H7: β = 0.359, p < 0.05, t = 2.451) and mediated the relationship between ETQ and RI (Hair et al., 2017).

Apart from the path coefficient and R² results of the structural model, the f² effect size explained the presence of mediation (Nitzl et al., 2016), and the Q² showed the predictive accuracy of the structural model of the structure (Hair et al., 2019). The f² values of 0.02, 0.15 and 0.35, respectively, were considered small, medium or large, and the value less than 0.02 indicates no effect (Cohen, 2013; Hair et al., 2019), and according to the results of f² (Table 4), only the effect size for H3 ETQ -> SAT (f² = 0.229) was medium. The remaining hypotheses were considered to have a small effect size. Finally, we determined the predictive correlation of the PLS path model by performing a blindfolding procedure using an omission distance D = 7 (Sarstedt et al., 2017). The results showed that the Q² value of SAT (0.136) and RI (0.375) were more than 0 for all endogenous constructs, indicating that the PLS-path model is correlated (Hair et al., 2019).

The study uses ETPB as a framework to examine ETQ and re-examine behavioural intention. According to the study of Jin et al. (2013), we put ETQ into a higher-order structure consisting of four sub-dimensions: GQ, IQ, OQ and PEQ. It found that many studies have adopted TPB to prove the direct relationship between SN, ATT, PBC and behavioural intention (Erul et al., 2020; Su et al., 2021). Some studies (He & Luo, 2020; Meng & Cui, 2020; Vesci & Botti, 2019) find that behavioural intention can be explained by all TPB factors (i.e. ATT, SN and PBC). The study’s findings support the earlier conclusions, suggesting that TPB structure predicts RI (Abbasi et al., 2020, 2021; Liu et al., 2021). By integrating the ETQ and tourists’ SAT, this study further enhances the current understanding of tourists’ SAT (Chen et al., 2023) and RI (Jeong et al., 2019).

The study also reveals how independent variables and mediations in the ETPB structural model affect RI. The results of this study find that the total effect, direct effect and indirect effect of ETQ and RI are significant, indicating a partial intermediary relationship between SAT with ETQ and RI. It is also verified that the tourist SAT is positively correlated with ETQ (Jeong & Kim, 2020; Jeong et al., 2019); the tourist SAT has a significant impact on RI (Hutchinson et al., 2009; Tang et al., 2022) and the positive impact of ETQ on RI (Ko et al., 2023). The results also showed that the main implications of ETQ on RI are direct, which recognises the intensity of the effects of ETQ on RI and determines the critical relationship between them. In the meantime, the SN are found to have the strongest influence on RI, and there is a positive correlation between them, which again confirms the previous study (Ko et al., 2023; Soliman, 2021; Tajeddini et al., 2021), which also supports the SN, that is, the perceived behaviour control is a prerequisite for intention and behaviour (Ajzen, 1991). Also, it explains that SN can affect individuals’ behavioural intentions mainly through familiar crowd suggestions, increasing the endurance of visitors to revisit their intention (Lam & Hsu, 2006). Anyway, the ETPB, which includes ETQ and SAT as newly added variables, is a helpful framework to convey a greater degree of ETQ effect to RI. The findings enhance our understanding of the importance of the correlation between ETQ and RI by developing strategies to promote visitors’ RI.

This study may have some practical significance for promoting tourists’ behavioural intention through the programme of improving ETQ to plan for the development of sports tourism. The ETQ measurement scale can be extended to the theoretical structure of TPB as an independent variable. Tourist destinations provide tourists with unique projects and special activities (Lam & Hsu, 2006). With high-quality events, tourists’ intention to revisit can be improved (He & Luo, 2020), while the cost of attracting tourists can be reduced (Kim et al., 2013). Organisers, local governments and managers of sports tourism events should fully consider the impact of ETQ on tourism development. Firstly, organisers can enrich the content of events by attracting talented and capable players to participate in various forms to improve the GQ of the events (Jeong et al., 2019). For example, organisers can provide useful information such as stadium location, match schedule, team information, hotels, restaurants and transportation (Jin et al., 2013) to attract the interest and attention of sports tourists because watching high-level matches and cheering for the favorite players are the motivation of tourists to attend sports events.

Secondly, the IQ between tourists and others should be improved to enrich cultural entertainment, artistic performances and interesting activities outside the event so that tourists can release their emotions in sports events and effectively interact with service providers, volunteers and other fans, to achieve the purpose of pleasing the mind and body, and truly obtain novel and unique experiences during the trip. It can increase the experience and SAT of visitors to the event, improve the possibility of visitors revisiting the destination and recommend it to family, friends and acquaintances. Then, tourists’ views on the OQ can be improved so that they can find their common topics and concerns when watching the events and communicating with fans, promote emotional resonance among tourists and satisfy the psychological needs of consumers in the consumption process, to improve the recognition and attention of tourists to sports tourism events. Finally, to enhance the tourists’ evaluation of the stadium atmosphere, safety and other features, the organisers should improve the related supporting hardware facilities of the stadium, should take the personal safety and property safety of tourists as a priority, and provide tourists with a high-quality event environment.